CSA
Summary
Read the report at California State Auditor ↗
Table of Contents
Summary S-1
Introduction 1
Chapter 1
Some County Treasurers Assume
More Risk Than Appropriate for Short-Term
Local Government Investment Pools 7
Chapter 2
Current Regulations and Guidelines Fail To
Address Investments Made By Agents 29
Chapter 3
Recommendations 41
Appendix A
Interest Allocations That Deviate
From Government Code Requirements 45
Appendix B
Selected Comparable Financial Data
of County Investment Portfolios for 1994 47
Appendix C
Counties’ Investment Pool Participants
by Type and Average Percentage of the Pool 49
Responses to the Audit
County of Colusa 51
California State Auditor’s Comments on
the Response by the County of Colusa 55
41
County of Monterey 57
California State Auditor’s Comments on
the Response by the County of Monterey 89
County of Placer 93
California State Auditor’s Comments on
the Response by the County of Placer 161
County of Sacramento 165
California State Auditor’s Comments on
the Response by the County of Sacramento179
County of San Bernadino 181
California State Auditor’s Comments
on the Response by the
County of San Bernardino 183
County of San Diego 185
California State Auditor’s Comments on
the Response by the County of San Diego191
County of Solano 195
California State Auditor’s Comments on
the Response by the County of Solano 199
County of Sonoma 203
California State Auditor’s Comments on
the Response by the County of Sonoma 227
Summary
42
C
ounty treasurers are responsible for the receipt and safekeeping
of all funds belonging to their respective counties and any
additional monies deposited by other participants in the counties’
Audit Highlights... investment pools. However, we found that several county
treasurers may employ risky investment strategies in their
Several counties management of public funds. These strategies run counter to
employ at least one of the counties’ investment policies and the principles of prudent
three investment investing, namely safety, liquidity, and yield, in that order. We
strategies we consider
surveyed 57 county treasurers and found that several counties,
too risky.
including 7 of the 8 counties we visited, employ at least one of
the three strategies that we believe can put public funds at risk.
These high-risk
Such high-risk strategies include holding excessive
strategies include
concentrations of often volatile structured notes, excessively
excessive
concentrations of leveraging or borrowing against portfolios through reverse
structured notes, repurchase agreements, and investing significant proportions of
overuse of reverse portfolios in securities with long-term maturities. These
repurchase strategies increase the affected portfolios’ sensitivity to interest
agreements, and rate changes and reduce the ability of county treasurers to meet
holding too many
unanticipated cash needs. Also, they expose portfolio
long-term securities.
participants to increased risks.
Additionally, three
Although the risky investment strategies we discuss in Chapter 1
counties utilized agents
were also used in Orange County, fortunately, none of these
who are also
practices were taken to the extremes we found in that county.
custodians to make
investments that are While we found that all counties we visited experienced
not required to be unrealized losses to their portfolios as of March 31, 1995, these
recorded in accounting losses would not be realized unless the securities whose values
records or disclosed to had declined were sold prior to maturity. The eight treasurers
participants. from these counties stated that they had sufficient liquidity in
their portfolios to preclude the sale of securities prior to maturity
and, therefore, none believe that they will incur any actual
losses.
Specifically, we found the following:
Seven of eight counties we visited held concentrations of structured notes in excess of 30
percent of their respective portfolios as of March 31, 1995. Many structured notes are not
actively traded and, therefore, are harder to accurately price and sell, are typically considered
less easy to convert to cash, and are more sensitive to changes in interest rates that can
decrease their market value;
Four counties we reviewed leveraged their portfolios by more than 40 percent sometime
during 1994, with one county leveraging its portfolio as much as 80 percent. Significant use
of leverage dramatically magnifies the portfolio’s exposure to risk; and
43
Six of the eight counties we reviewed managed portfolios of investments with maturities
averaging 2.5 years or more during 1994. The average maturity for one county’s
investments was an astounding 27.9 years. When most of a portfolio is made up of
securities with long maturities, the portfolio is much more sensitive to interest rate changes,
and the treasurer’s ability to meet unforeseen cash demands is more limited.
According to our investment experts, employing these investment strategies can be inappropriate
for short-term investment pools like those used by the counties because they do not emphasize
safety and liquidity over yield and expose the portfolios and their participants to increased risks.
Furthermore, some counties are using agents, who in many cases also act as the securities
custodians for the same counties, to execute securities lending and/or reverse repurchase
transactions on behalf of the counties. Agents acting on behalf of three counties we reviewed
performed hundreds of millions of dollars worth of securities lending and reverse repurchase
transactions that current regulations and guidelines fail to address. Therefore, these transactions
are not reported to oversight agencies and pool participants, were not reflected in the counties’
financial statements or clearly disclosed in footnotes, and were not considered by the counties
when determining compliance with pertinent investment laws and county policies.
In addition, we found the following:
The investment pools of the three counties that delegate their investment authority to agents
bear all the risk if investment losses occur, while the agents share between 30 and 50 percent
of any profits. In one county these rates were not competitively bid; and
Two of the three agent agreements we reviewed allowed the agents to buy and sell securities
with their affiliates. This allowance raises questions about ethical and prudent investments.
Recommendations
To improve the investment practices of local governments, we recommend that the Legislature
amend the California Government Code. A few of our key recommendations are to:
Require written investment policies for all local governing bodies to ensure that safety and
liquidity are paramount to yield;
Limit the use of reverse repurchase agreements to 20 percent of the portfolio and only for
specified purposes;
Establish and define a prudent person rule for local investment officers;
Limit the use of derivatives or other structured investment instruments and prohibit those that
put principal at risk. None of these instruments are to be purchased with borrowed or
leveraged funds. Further, the derivatives or structured investments purchased should be
openly traded in the secondary market on a recognized exchange. Any investments in these
instruments should be limited to no more than 5 percent of the portfolio;
44
Competitively bid separately for lending agent and custodial services, and ensure that all
county investment activity by agents or the county are properly recorded and disclosed to
interested parties; and
Require investment reports, at least quarterly, to the governing body and investment
participants.
To manage county investment portfolios using the basic principles of prudent investment
practices, we recommend that the treasurers of the counties that employ any of the high-risk
strategies do the following:
Avoid the use of these risky strategies in the future;
Prudently divest themselves of structured notes that add risk to their respective portfolios;
Limit future purchases of structured notes that increase risk to the portfolio (as measured by a
duration analysis) to no more than 5 percent of all investments;
Confine the use of leverage in their respective portfolios to 20 percent or less; and
Prudently restructure their respective portfolios to reach an average maturity of no more than
2.4 years.
Further, those county treasurers who participate in a security lending program should do
the following:
Include agent transactions in statutory limitations and compliance reviews;
Report on the security lending program in the counties’ annual financial reports and in
reports to boards of supervisors and pool participants;
Competitively bid for lending agent and custodian services; and
Preclude agents from dealing with their affiliates.
Agency Comments
Because the counties believe we used Morningstar Mutual Funds as the basis for concluding that
their respective portfolios were too risky, they disagreed with many of the findings and
recommendations contained in our report. Specifically, the counties disagreed with our
characterization of the counties’ use of structured notes and the average length of maturity for
their respective portfolios. Individual counties also raised other concerns relating to the tone
and content of the report. Since we did not conclude on the counties’ investment practices
based on the Morningstar data as they assert, we have responded to these and other concerns
raised by counties after each county’s response.
45
Introduction
E
ach of California's 58 counties has a county treasurer who is responsible for investing
funds for the county and other entities, such as school districts and special districts that
participate in the county’s investment pool. The California Government Code specifies
the types of financial instruments treasurers may invest in. Fifty-seven county treasurers
reported managing investments totaling $29.4 billion as of June 30, 1994.1 In addition, they
reported an average yield of 4.47 percent and an average maturity of 1.4 years for their
investments.
Treasurer’s Responsibilities
According to the California Government Code, Section 27000, the county treasurer is
responsible for receiving and keeping safe all monies belonging to the county and all other
monies directed by law to be paid into the treasury. The code generally requires that the county
treasurer hold all monies of the county and the school districts. Also, other public agencies,
such as cities and certain special districts, can voluntarily deposit their monies into the county’s
treasury.
Furthermore, Section 27100.1 of the code specifies a trust relationship exists between the
treasurer and public entities or public officials who are acting in a fiduciary capacity when they
deposit funds into the county treasury. As the trustee, the treasurer should make investments
with care, skill, prudence, and diligence.
Most counties have investment policies stating that investments should be made following the
fundamental principles of safety, liquidity, and yield—in that order. While the county treasurers
have a trust responsibility to the investment pool participants, the California Government Code,
Sections 53601 and 53635, allows county treasurers great latitude in their investment practices.
Specifically, the law allows the purchase of securities such as the following:
U.S. Treasury notes and bonds;
Bonds and notes of any local agency within California;
Bankers’ acceptances—bills of exchange or time drafts, drawn on and accepted by a
commercial bank, that are eligible for purchase by the Federal Reserve System;
Prime quality commercial paper—a debt security with a maximum maturity of 180 days;
Negotiable certificates of deposit;
1 We reported on Orange County’s investment practices in our report entitled Orange County:
Treasurer’s Investment Strategy Was Excessively Risky and Violated the Public Trust.
46
Repurchase and reverse repurchase agreements; and
Mortgage-backed securities, which include government agency securities, such as Federal
Home Loan Bank (FHLB), Federal National Mortgage Association (FNMA), other U.S.
government agency issues, and miscellaneous mortgage-backed securities.
Finally, the California Government Code, Section 53684(b), requires county treasurers to
apportion quarterly to the local agencies or districts any interest earned from the investment of
funds. The interest earned should be in an amount proportionate to the average daily balance of
the amounts the local agencies or districts deposit.
Size, Yield, and Average Maturity
of the Counties’ Investment Portfolios
The 57 county treasurers managed a total of approximately $29.4 billion in investment funds as
of June 30, 1994. The net investments at their purchased cost (book value) ranged from a low in
Sierra County of $6.7 million to a high in Los Angeles County of $6.9 billion. Appendix B
presents selected comparative financial data of the 57 county investment portfolios. The data
includes the book and market values of each portfolio, the average annual yield, average
portfolio maturity, average percentage of structured notes of each portfolio, and
average percentage of portfolio leveraged. We classified the counties’ portfolios into the
following four categories based on the amount of investments being managed as of June 30,
1994:
Table 1
Percentage of Statewide Investment
Portfolio Based on County Portfolio Size
Number Percentage of
of Percentage of Statewide
Size of Portfolio Counties Total Countiesa Portfolio
Small - under $100 million 24 42% 3%
Medium - $101 million to
$500 million 19 33 14
Large - $501 million to
$1 billion 6 11 16
Very Large - more than
$1 billion 8 14 67
Total 57 100% 100%
a Excluding Orange County.
The average yields (rate of earnings returned to participants from their deposits) during calendar
year 1994 ranged from a low in San Mateo County of 3.07 percent to a high in Colusa County of
7.32 percent. The average yield was 4.47 percent for all 57 counties. (See Appendix B for
more details.)
47
The average portfolio’s maturity (the average length of time securities need to be held before
they mature) was 1.38 years. Twenty-seven of the counties reported average maturities of less
than one year. Four counties (Colusa, Lassen, Placer, and San Diego) had average maturities of
more than three years. (See Appendix B for more details.)
Participants in the Counties’
Investment Pools
As shown in Table 2, we compiled the counties’ data on the types of entities participating in the
counties’ investment pools. Approximately 56 percent of the funds on deposit in the county
pools during 1994 represented funds under the control of the various county boards of
supervisors. Approximately 30 percent of the funds represented deposits by school districts.
State law requires school districts to deposit their funds in the county treasury unless the school
board declares funds to be excess funds, in which case funds can be invested outside the county
treasury. Approximately 5 percent of the funds represented deposits by special districts, such as
cemetery districts, that are required to deposit their funds in the county treasury. An additional
5 percent of the funds represented deposits by special districts, such as fire districts, that may
appoint a treasurer other than the county treasurer. Other public agencies, such as cities,
community college districts, and various other entities that do not report to the county board of
supervisors, voluntarily deposit their funds in the county pools and comprise approximately
4 percent of the pools. (For additional detail, refer to Appendix C.)
Table 2
Number and Types of Participants
Depositing Funds in County Pools
Percentage
Number of of Total
Participants Participants
County controlled funds N/A 56%
School districts 1,034 30
Involuntary special districts 1,474 5
Voluntary special districts 923 5
Various other voluntary
participants 582 4
100%
Scope and Methodology
The California Legislature requested the Bureau of State Audits to determine the prevalence of
risky borrowings and investment practices for 57 of the 58 counties within California. This
request followed Orange County’s December 1994 bankruptcy filing, which occurred as a result
of risky investment practices by the Orange County treasurer. We excluded Orange County
from our review of California’s counties as a whole because we separately audited its investment
practices. We issued our report, Orange County: Treasurer's Investment Strategy Was
Excessively Risky and Violated the Public Trust on March 28, 1995.
48
We surveyed the other 57 counties to identify various investment practices and to obtain other
pertinent information. Furthermore, we visited 8 counties (Colusa, Monterey, Placer,
Sacramento, San Bernardino, San Diego, Solano, and Sonoma) based on our preliminary
assessment of survey information on the risk and diversity of each county’s investment portfolio.
During the audit, we reviewed the following for the eight counties:
The investment strategies of the eight county treasurers;
The degree of risk in the portfolios and an analysis of the treasurers’ investments;
The counties’ internal control structures and investment disclosure policies as they relate to
the counties’ investment practices; and
The treasurers’ trust responsibilities to act for the benefit of the counties and other pool
participants, including allocating interest income and treating all participants consistently.
To assist us in evaluating the investment strategies of the eight selected
counties, we engaged the services of an investment consulting firm—Analysis Group, Inc.
These investment experts performed a review of the counties’ investment strategies as of
September 30, 1994, and March 31, 1995, including an assessment of the degree of investment
risk in the portfolios and an analysis of the eight counties’ investments.
Also, we reviewed the eight counties’ records to determine whether the treasurers’ or county
auditor-controllers’ staff used appropriate methods to allocate interest earnings to participants.
Appendix A provides the results of this review.
Finally, we mailed two questionnaires about local government investment practices to 57
counties, excluding Orange County, to compile comparative investment-related information for
these counties. When necessary, we followed up on these questionnaires to ensure complete
information from the county treasurers. Appendices B and C provide selected comparative
information for the 57 counties.
49
Blank page inserted for reproduction purposes only.
50
Chapter 1
Some County Treasurers Assume More
Risk Than Appropriate for Short-Term
Local Government Investment Pools
Chapter Summary
C
ounty treasurers are responsible for prudently managing public funds in their custody.
According to a survey we conducted of 57 county treasurers, several provided indications
that they may employ, to some degree, at least one of the three investment strategies we
consider risky for short-term portfolios. These strategies included purchasing large
concentrations of structured notes, excessively leveraging or borrowing against their portfolios,
and investing significant proportions of their county portfolios in longer-term securities, all in an
effort to achieve greater returns. Seven of eight counties we visited used one or more of these
investment strategies in a manner inconsistent with their investment practices, thus exposing
their portfolios to additional risk.
Although the risky investment strategies we discuss in this chapter were also used in Orange
County, fortunately, none of these practices were taken to the extremes we found in that county.
While we found that all counties we visited experienced unrealized losses to their portfolios as of
March 31, 1995, these losses would not be realized unless the securities whose values had
declined were sold prior to maturity. The eight treasurers from these counties stated that they
had sufficient liquidity in their portfolios to preclude the sale of securities prior to maturity and,
therefore, none believe that they will incur any actual losses.
Basic Principles of
Prudent Investment Practices
When funds are invested on behalf of public entities, the fundamental principles are safety,
liquidity, and yield—in that order. Safety, the preservation of investment capital, is the
foremost objective. To ensure the preservation of capital, the investor should restrict the
investments to high-quality, low-risk securities purchased from reputable dealers. Liquidity, the
second priority, is the ability to readily convert investments to cash to meet the spending needs
of the investment pool participants. To ensure liquidity, investments should be limited to
short-term securities that are actively traded in a secondary market. A secondary market is one
in which there are sufficient buyers and sellers, so that a particular asset can be readily converted
to cash. In addition, more liquid investments typically have a smaller spread between the price
offered by the buyer and the price asked by the seller. Yield, the return on investments, is the
third priority. It should not become a consideration until the basic requirements of safety and
liquidity have been met.
51
Some Counties Are Using
Risky Investment Strategies
As part of our survey, we requested that each of the 57 counties send us a copy of its respective
investment policy in effect for calendar year 1994. Twenty-two counties indicated they had
changed their policies in, or after, December 1994. Most
counties’ investment policies included the fundamental
principles of safety, liquidity, and yield in the same order of
Some county treasurers
priority as described earlier. Using the results of our survey of
are using investment
57 county treasurers, we selected eight counties for review
strategies that do not
seem to follow their stated (Colusa, Monterey, Placer, Sacramento, San Bernardino,
policies. San Diego, Solano, and Sonoma). Based on our survey and
review of eight counties, some county treasurers are using
investment strategies that do not seem to follow their stated
policies. The results indicate that several counties employ one
or more of the following strategies:
Investments in Structured Notes—27 county treasurers indicated they invested a
proportion of their portfolios in structured notes during 1994. Depending upon the type of
structured notes, holding concentrations of these notes could add risk to the portfolio. Most
structured notes are derivative securities, in which the interest rate as well as other features
vary over the life of the debt. Derivative securities are securities whose value is based on, or
derived from, some underlying asset, reference rate, or index. Structured notes are typically
less liquid than fixed rate notes because of their custom design. The custom design can also
make them more sensitive to changes in interest rates than fixed rate notes. These
conditions can make structured notes more difficult to price and reduce their value in the
secondary market. Further, depending on the make up of the structured notes, some may
add increased risk to the portfolio;
Using Reverse Repurchases for Leverage—5 county treasurers reported borrowing, termed
leverage, against more than 20 percent of their respective portfolios. Significant use of
leverage can dramatically increase the portfolio’s exposure to risk; and
Holding Investments With Lengthy Maturities—10 county treasurers reported that their
portfolios had average weighted maturities during 1994 of greater than 2.4 years. Holding
large concentrations of long-term securities increases the portfolio’s sensitivity to interest
rate changes and may reduce the investment fund’s ability to meet unanticipated cash needs.
Comparative Investment Strategies
To evaluate the investment strategies used in county portfolios, our investment experts compared
the investment policies of the eight counties we selected for review to the investment strategies
used by portfolios with similar objectives and constraints. In attempting to identify comparable
portfolios, our experts initially reviewed over 2,800 fixed income mutual funds tracked by
Morningstar Mutual Funds including extremely short-term money market funds and various
categories of bond funds operated by mutual fund companies. These bond funds ranged from
those with investment criteria requiring high-quality, short-term investments to those funds
allowing investments in low-quality, long-term securities. Our investment experts determined
52
that high-quality, short-term bond funds were the most comparable to local investment portfolios
for the following reasons.
First, these bond funds require high-quality investments similar to those prescribed by the
California Government Code as appropriate for investment by county treasurers. Second, these
bond funds limit the length of maturity for investments just as the Code limits the length of
maturity for investments made by county treasurers. County treasurers may be granted specific
approval by their respective board of supervisors to invest in longer term securities. Third,
although these bond funds are comprised of voluntary depositors, the funds are able to forecast
their anticipated cash flows much like county treasurers, allowing the bond funds to invest a
portion of their portfolios in securities with maturities greater than one year. For example, while
the maturity of these bond funds made up of voluntary depositors averages 2.4 years, the average
portfolio maturity reported by the 57 county treasurers we surveyed was significantly shorter at
1.4 years, even though the majority of the funds deposited with county treasuries cannot be
withdrawn and invested elsewhere.
Furthermore, our investment experts believe that the average
maturities of these bond funds are not overly short in providing
comparative investment criteria to county portfolios. Counties
Comparing high-quality, return a dollar for every dollar deposited when needed by pool
short-term bond funds to participants to pay operating costs, unlike these bond funds that
county investment
return the proportionate market value of the deposits in the fund
portfolios is reasonable
when a depositor requests a withdrawal. Therefore, bond funds
and not overly
have more flexibility to invest in longer term securities than do
conservative.
county portfolios. Based on all these factors, we feel that the
selection of high-quality, short-term bond funds as a comparison
group to county investment portfolios is reasonable and not
overly conservative.
Of the more than 2,800 mutual funds tracked by Morningstar Mutual Funds, 252 met the criteria
for the comparison group of high-quality, short-term bond funds. Thirty-three of these funds
voluntarily reported data concerning the average percentages of structured notes held by the
funds. Twenty-nine of the 33 funds (88 percent) reported holding no structured notes at all.
The remaining 4 funds reported holding concentrations of structured notes ranging from one
percent to 10.5 percent for an overall average of 5.9 percent. In addition, 36 funds voluntarily
provided data regarding duration (a measurement of a fund’s sensitivity to changes in interest
rates) which averaged 1.8. Finally, all 252 funds reported on the maturities for their respective
funds which averaged 2.4 years. All data reported by the comparison group bond funds covered
the period from October 14, 1994, through March 17, 1995.
Investments In Structured Notes
As stated above, our investment experts identified portfolios with investment styles and
objectives comparable to local government short-term pools. Using this data, our experts
determined that portfolios with comparable investment styles and objectives to the county
portfolios held average concentrations of structured notes equal to approximately 5.9 percent of
the respective portfolios. Depending on the type of structured notes held, holding
concentrations of structured notes in excess of 5.9 percent may signal an increased risk that the
portfolios are more sensitive to interest rate changes than similar portfolios holding smaller
53
proportions of such notes. Of the 27 counties who reported holding structured notes in their
portfolios, we selected 7 for further review. The 7 counties we reviewed had portfolios
containing from 33 percent to 72 percent of structured notes as of March 31, 1995. Figure 1
shows the average percentage of structured notes held during 1994 by the 57 counties we
surveyed.
Figure 1
Structured Securities Held by Counties
6% - 9.99% 10% - 24.99%
(9 c ounties ) (6 c ounties )
G reater than
25%
(8 c ounties )
Les s than 6%
(34 c ounties )
Percent Held By 57 Counties
Structured notes are investments that are generally derivative
securities in which the interest rate, as well as other features, varies
over the life of the debt. Derivative securities are financial
Depending upon the instruments whose value is based on, or derived from, some
composition of the underlying asset, reference rate, or index. Structured notes are
structured notes, they may “structured” with specific features to fit the needs and preferences
significantly increase or of particular investors and, depending upon the note’s composition,
decrease the overall risk
may significantly increase or decrease the overall risk in the
in the investment portfolio.
investor’s portfolio. For example, structured notes can be
designed to reflect the investor’s opinion on the future course of
interest rates, or other financial variables. In other words,
structured notes allow an investor to make a “bet” on future
financial events. For example, if an investor believes interest rates
will fall, a structured note called an inverse floater can be designed so that the return on the
investment would rise if interest rates fell. If, however, the investor’s expectation is wrong and
interest rates rise instead of fall, the value of the investment and the return on the investment will
decrease. Other structured notes can be tied to move in the same direction as interest rates or
other underlying indices. Typically, these floating rate notes are less volatile than inverse
floaters. Because structured notes can be designed to have any level of risk, they can
significantly increase or even decrease risk to a portfolio resulting from interest rate changes.
54
As stated in the Introduction, the California Government Code, Sections 53601 and 53635,
allows local government investment officers great latitude in their investment practices.
Specifically, the law allows the purchase of such securities as U.S. Treasury notes and bonds,
bankers’ acceptances, prime quality commercial paper, reverse repurchase agreements, and
mortgage-backed securities. The inherent risks associated with these investments range from
low-risk U.S. Treasury securities to high-risk notes structured from allowable securities.
One way to classify securities is according to the structure of the note’s coupon payment. A
coupon is the periodic interest payment made to the holder of the note during the life of the
security. The predominant issuers of structured notes include the Student Loan Marketing
Association (SLMA), the Federal Home Loan Bank (FHLB), the Federal Farm Credit Bank
(FFCB), the Federal Home Loan Mortgage Corporation (FHLMC), the Federal National
Mortgage Association (FNMA), corporate issuers, and issuers of foreign and domestic
negotiable CDs. Our investment experts analyzed the portfolios of the eight counties we
reviewed, and identified the following coupon structure categories of notes held in the portfolios:
Fixed rate notes: These have a fixed coupon payment for the life, or term, of the security.
Floating rate notes: These have coupon payments that rise and fall as a particular interest
rate index increases or decreases.
Inverse floating rate notes: These have a portion of their coupon payment that falls as
interest rates rise, including notes whose principal payments fall as interest rates rise.
Step notes: These have coupon rates that either increase (step up) or decrease (step down),
depending on the type of step note.
Indexed call notes: These are notes where the redemption period is dependent on some
index (also called trigger notes).
After their analysis, our investment experts stated that the strategy used by six of seven counties
holding high proportions of thinly traded structured notes in their respective portfolios can
substantially increase the sensitivity to changes in the interest rates and may impair liquidity.
Figure 2 illustrates the types and percentages of structured notes held as of March 31, 1995 by
the seven counties we reviewed.
Figure 2
Types and Percentages of Structured
Notes Held by Counties We Reviewed
As of March 31, 1995
55
Sonoma
Solano Index Calls
Step
San Diego Floating
Inv. Floating
San Bernardino
Placer
Monterey
Colusa
0 20 40 60 80
Percent
Percent of total portfolio made up of various structured notes.
In Colusa County, for example, an analysis of the coupon structure of the portfolio indicates that
it is comprised of nonstructured fixed rate notes (30 percent), structured floating rate notes (22
percent), and structured inverse floating rate notes (48 percent). In fact, the majority of the
county’s portfolio (70 percent) is comprised of thinly traded structured floating notes as of
March 31, 1995. Thinly traded notes are those not actively traded in a secondary market.
Our investment experts also calculated the duration for each county. Duration is a quantitative
measure of a portfolio’s sensitivity to a one percent interest rate change. A portfolio that does
not put principal at risk as interest rates fluctuate, like a passbook savings account, has a duration
of essentially zero.
However, as investments are made in longer term securities, and structured notes are added to a
portfolio, the portfolio duration increases to reflect increased sensitivity to interest rate changes.
According to our investment experts, the average duration comparable for short-term local
investment pools is 1.8. However, the duration for Colusa County’s portfolio was much higher,
measuring 22 as of March 31, 1995, excluding the effect of leverage. This high duration of 22
means that if interest rates were to increase by one percent on March 31, 1995, the market value
of Colusa County’s portfolio would decrease by approximately $4.3 million on a portfolio
market value of $19.4 million. Compared to the duration of 2.7
computed on the portion of its portfolio invested in
nonstructured, fixed rate notes, Colusa’s structured notes with a
As investments are made
in longer term securities duration of 30.3 were more than 11 times as risky. In other
and volatile structured words, Colusa County’s structured notes were 11 times as risky
notes, a portfolio’s as its fixed rate notes in terms of interest rate risk. Interest rate
sensitivity to interest rate risk, sometimes called market risk, is the risk that the value of a
changes fixed income security will change with changes in interest rates.
increases. In general, the market value of a fixed income security will move
56
in the opposite direction of a change in interest rates. For example, if interest rates rise, the
market value of a fixed income security will fall. This decline in market value would be
realized as a loss only if the security had to be sold before it matured. If the security is held to
maturity it will pay its full face value.
Placer County is another example of a county holding excessive concentrations of structured
notes. Almost three quarters (72 percent) of Placer County’s portfolio was comprised of
structured notes as of March 31, 1995. The proportion of inverse floaters was 52 percent of the
portfolio as of September 30, 1994. Six months later, inverse floaters represented 64 percent of
the portfolio. However, the proportional increase in inverse floaters was caused by a decrease in
the overall size of Placer County’s portfolio. According to our investment experts, investing in
such high concentrations of inverse floaters indicates the treasurer expects interest rates to fall,
yet the investment exposes the portfolio to the risk that they may rise. For example,
Placer County’s portfolio duration as of March 31, 1995, was 3.4. A portfolio duration of 3.4
would cause a decline of approximately $12.2 million in the market value of Placer’s $357.7
million portfolio, excluding any leverage effect, if interest rates increased by one percent on
March 31, 1995. Further, the duration of 4.2 on Placer County’s structured notes was 3.5 times
higher than the duration computed for its nonstructured fixed rate notes of 1.2, meaning that the
interest rate risk for its structured notes was 3.5 times higher than its risk for nonstructured fixed
rate notes.
In contrast, San Bernardino County invested over 30 percent2 of its portfolio in structured
floating notes as a hedge against rising interest rates. Floating
notes with interest rates that float are beneficial when interest
rates climb since the coupon rate would increase as the
One county’s investment
underlying interest rate index rose. However, if rates decline,
of over 30 percent in
the market value of the floating notes do not increase as much as
structured floating notes
comparable fixed rate notes.
actually decreased the
overall risk to the portfolio.
The risk related to structured notes is dependent upon the
In contrast, investments in features embedded in the notes and lengths of maturity of the
inverse floaters by two instruments. In reviewing the portfolios of Colusa, Placer, and
counties exposed their San Bernardino counties, we found that the risk as illustrated by
portfolios to significant duration varied widely between the types of structured notes.
interest rate risk. For example, the high risk of inverse floaters is demonstrated by
the 4.5 duration computed for Placer’s holdings of these notes at
March 31, 1995, and, to a more significant extent, Colusa’s
investment of 48 percent of its portfolio in inverse floaters at
March 31, 1995, with a huge 42.0 duration as of that date.
Therefore, when used in concert with a portfolio’s other investments, structured notes can either
reduce or increase the sensitivity of a portfolio to changes in interest rates. For example, the
duration computed by our investment experts as of March 31, 1995, for San Bernardino County’s
floating rate notes of .7 was less than the duration of its nonstructured fixed rate notes of 1.7.
Therefore, the combination of these investments resulted in San Bernardino’s reasonable overall
portfolio duration of only 1.4. Although Placer County used fixed rate and a small proportion of
structured floating rate notes to reduce the risk of its high concentration of inverse floaters, the
2 Approximately 10 percent of the county’s portfolio represents its investments in mutual funds that are
comprised of floating rate notes.
57
4.2 duration for its structured notes is in our opinion too high, with its structured notes 3.5 times
riskier than its fixed rate notes. Furthermore, in Colusa County it is clear that its selection of
floating rate notes with a duration of 5.0 significantly heightened, rather than decreased, the
interest rate sensitivity of its portfolio.
Overuse of Reverse Repurchases
for Leverage
Five of the counties we surveyed reported leveraging their investment portfolios through the use
of reverse repurchase agreements (reverse repos) in excess of 20
percent. We visited four of the five and found they reported
leverage ranging from 28.6 percent to 43.8 percent during 1994.
Most investment
The Government Finance Officers Association (GFOA), the State
authorities agree that
Treasurer’s Office, Standard & Poor’s, and the State Taskforce
leveraging should be
on Local and State Investment Practices, all recommend that
limited to no more than
20 percent of the portfolios containing public funds should limit leverage
portfolio. (borrowing) to a small percentage of the portfolio (for example,
10 to 20 percent). While county treasurers are not required to
follow the recommendations of any of these groups, we think a
20 percent limitation on leveraging is appropriate for local
government investment funds.
Investment in reverse repos can increase the yield to a portfolio, although not without assuming
additional risk associated with these agreements. In a reverse repo, the owner of a security, such
as a county, “borrows” by selling the security to an investment broker with an agreement to
repurchase it a short time later. In effect, the security held by the broker is collateral for a loan.
In a reverse repo transaction, the county agrees to pay a stipulated rate of interest to the broker as
the cost of borrowing the money. The county receives short-term cash from the broker, without
permanently relinquishing ownership of the underlying security. The county can then reinvest
the cash received, leveraging the original principal by, in effect, investing the same money twice.
If the spread (that is, the earnings on the investment minus the cost of the borrowing) is positive,
the reverse repo provides additional income to the county. The greater the spread, the higher
the yield to the portfolio.
The use of reverse repos is more risky when a county uses the proceeds to purchase securities
having variable interest rates that fluctuate in ways that do not match the ways the rates on the
reverse repo agreements fluctuate, the cash receipts from maturing securities are unequal to the
amounts of the reverse repos, or maturity dates do not match the termination dates of the reverse
repo agreements. Such arrangements add to the risk normally associated with reverse repos.
Figure 3 indicates those counties we reviewed that leveraged their portfolios by more than 20
percent during 1994, and the figure also shows the degree of leverage as of September 30, 1994,
and March 31, 1995.
Figure 3
Percentage of Leverage
Through Reverse Repurchases
58
Of the four
100%
counties we
reviewed that used 90% Sept. 30, 1994
reverse repos in 80% March 31, 1995
excess of 20
70%
percent, all failed
60%
to match the
maturity dates of 50%
the securities 40%
purchased with
30%
borrowed funds to
20%
the settlement dates
of the respective 10%
reverse repos. In 0%
other words, all Monterey Placer San Bernardino Solano
four counties were
As a Percentage of the Base Portfolio
buying long-term
securities with short-term borrowing.
When counties do not match the maturity dates of securities
purchased with borrowed funds to the settlement dates of the
Of the four counties we underlying reverse repos, the risk to the counties’ portfolios
reviewed, none matched increases. For example, reverse repo borrowing rates are usually
the maturity dates of the guaranteed only for short periods (for example, 180 days or less).
securities purchased with When the return rate of the security purchased using the
borrowed funds to the borrowed funds is for much longer (for example, two to four
settlement dates of the
years), the county must continually borrow at the then current
respective reverse
interest rates until the purchased security matures. Increases in
repos.
short-term interest rates translate into higher reverse repo
borrowing costs, thus reducing, or even eliminating, the spread
between the borrowing costs and investment returns.
Solano County reported that an average of 40 percent of its portfolio was leveraged through
reverse repos during calendar year 1994. The maturity dates of
the securities purchased with borrowed funds extended two to
five years while the settlement dates of the related reverse repos
One county initially
were for 30 to 180 days. The assistant treasurer stated that
matched maturity dates on
initially the maturity dates of the securities purchased with
its reverse repos, but
borrowed funds did match the settlement dates of the related
began buying investments
reverse repos. However, the county later began buying longer
with longer maturities to
term securities with short-term borrowings to take advantage of
achieve higher yields.
the higher rate of return available on longer term investments.
As interest rates began rising, the county made a concerted
effort to reduce the amount of leverage within the portfolio.
Between September 30, 1994, and March 31, 1995, reverse repos were reduced from 54 percent
to 32 percent of the portfolio.
59
Our investment experts computed an effective duration for Solano County as of March 31, 1995,
of 3.6 compared to the portfolio duration on that date of 2.7. Portfolio duration measures the
effect that the average maturity and the structure of the securities have on the portfolio.
Effective duration adjusts for the effect that leverage has on the portfolio. Therefore, using the
effective duration of 3.6, Solano County would incur a market loss of approximately $9.1 million
on the market value of its base portfolio of $254 million if interest rates were to increase by
one percent on March 31, 1995.
According to the assistant treasurer, Solano County has been rolling over, or renewing, a portion
of the remaining reverse repos it has rather than selling off the securities purchased with
borrowed funds and ending the agreements on the settlement dates. This is because under
prevailing market conditions, the county expects to lose less money by continuing to borrow and
pay the related costs than it would by selling at a loss all the securities purchased with the
borrowed funds. Solano’s assistant treasurer stated that the portfolio currently has sufficient
liquidity to keep rolling over these reverse repos and to continue to meet the county’s cash-flow
needs.
In February 1995, Solano County hired an independent
investment consulting firm to analyze the county’s portfolio.
The consultant’s report warned that leverage from reverse repos
The county’s own
exposed the county’s portfolio to increased market risk.
consultant concluded that,
Specifically, the consultant stated that by matching 30-day to
as interest rates rose, the
180-day reverse repos against securities maturing in two to five
costs associated with
borrowing exceeded the years, the county exposed the portfolio to the risk that security
earnings on the securities values would fluctuate based on changes in the level of interest
purchased. rates. Furthermore, Solano’s consultant stated that most of the
securities purchased with the borrowed funds had interest rates
that behave differently than the interest rates associated with the
underlying reverse repos. The consultant concluded that, as
short-term interest rates rose, the overall cost of borrowing rose. As a result, those costs
associated with borrowing money to purchase securities exceeded what the county was earning
on those securities. Solano’s consultant estimated that for every percentage point increase in
interest rates, the county’s annual net borrowing costs for its reverse repos would increase by
$1.26 million.
One County Uses a Line of Credit
for Reverse Repurchases
One county uses a different method for leveraging with reverse repos; however, the risk to the
county is the same as for counties using “traditional” reverse repos. In response to our initial
survey, the treasurer of Monterey County reported that the county was not engaged in traditional
reverse repos. Instead, according to the treasurer, he has an informal arrangement with a broker,
in which the broker extends the county an unsecured line of credit to provide the funds necessary
to purchase a security. According to this informal arrangement, the purchased security is used
as collateral for the loan; in effect, the security is reversed against itself. Further, future interest
payments on the purchased security are intended to provide sufficient funds to cover the cost of
interest charged on the line of credit and provide interest revenue to the county.
60
Monterey County’s treasurer maintains that by reversing a
security against itself, there is no risk to the county’s other
securities that were not purchased with borrowed funds (base
We disagree with one
county’s contention that portfolio). Although no portion of the base portfolio is initially
using a line of credit does pledged as collateral for the reverse repos described above, we
not expose its base disagree with the premise that the base portfolio could never be
portfolio to loss. impaired. If the market value of the securities serving as
collateral decline, the county would have to make up any shortfall
using available liquidity or, if sufficient liquidity was not
available, would need to sell or pledge some of the base
portfolio’s securities.
According to the treasurer, until mid-1994, Monterey County had used both traditional and
single-security reverse repos and pledged some of the base portfolio’s assets as collateral for the
traditional reverse repos. In mid-1994, the treasurer began using single-security reverse repos
exclusively and began accounting for the transactions in a separate portfolio. A short time later,
he began buying down the county’s traditional reverse repos and incorporating the securities into
the base portfolio. As of September 30, 1994, the par value of the county’s base portfolio was
$317.5 million while its total portfolio stood at $574.6 million. This 81 percent difference
results from investments purchased with borrowed money: either proceeds from traditional
reverse repos or from money available from the line of credit. By March 31, 1995, no portion of
the base portfolio provided collateral for borrowed funds; however, the total portfolio of $549.8
million was 63 percent larger than the county’s base portfolio of $337.9 million, the difference
representing securities purchased with borrowed funds. According to our investment experts,
the portfolio duration for Monterey County as of March 31, 1995, was 2.4 while its effective
duration was 3.9 for the same date, the increase indicating the effect of leverage on the portfolio.
Using the effective duration of 3.9, Monterey County would suffer a loss of approximately
$12.2 million on the market value of its base portfolio if interest rates were to increase by one
percent on March 31, 1995.
Collateral Calls Can Add to Risk
An added risk associated with reverse repos, particularly when the county is borrowing short and
buying long, is the exposure of the portfolio to an increased risk of collateral calls. Collateral
calls result when the assets pledged to secure a loan decline in value. When a broker lends
money on a reverse repo, the broker requires collateral in excess of the amount loaned to protect
the broker’s interests. If the market value of the collateral declines, the broker will require the
borrower to provide additional assets to restore the collateral to its original level. Collateral
calls can place untimely and significant demands on a portfolio by draining its available liquidity
or by requiring the premature liquidation of other assets. Ten of the 57 counties surveyed,
including 5 of the 8 we visited (San Diego, San Bernardino, Placer, Sonoma, and Solano
counties) reported receiving collateral calls from lenders during calendar year 1994. None of
the 10 counties who reported receiving collateral calls indicated they suffered any loss to the
principal in the base portfolio as a result of these calls. However, a collateral call requires that
these treasurers either pledge additional portfolio assets as collateral or contribute additional cash
to the dealer.
61
Lengthy Investment Maturities Can
Increase Interest Rate Risk
Ten county treasurers reported investing a significant proportion of their respective portfolios in
long-term securities during 1994, thus increasing their average weighted maturities to greater
than 2.4 years. We reviewed six of these ten counties. Long-term interest rates are typically
higher than short-term interest rates because a security with a longer maturity has greater interest
rate risk, therefore, the rate of return increases with the length of maturity to compensate
investors for this added risk. High concentrations of long-term securities can limit the
investment options available for some counties. This limitation occurs when long-term
securities with fixed coupon payments are purchased when interest rates are low and
subsequently interest rates rise, as was the case during 1994. During such a period of rising
rates, the sale of these long-term securities before their maturity would result in losses to the
counties’ portfolios.
In selecting longer term securities for investment, the county
accepts an increased risk (called market risk) that prevailing
High concentrations of interest rates may rise. Typically, in that event, the market value
long-term securities limit of the long-term security declines below its original purchase
investment options and price. This decline would be realized as a loss by the
may cause losses if investor if the security were sold before its maturity. If held to
treasurers must sell to maturity, the investment will return, or pay, its face value.
meet cash demands.
Because of the potential losses associated with selling securities
before their maturity, a county treasurer may be unable to
restructure the portfolio to meet unanticipated demands for cash
by investment pool participants.
Figure 4 shows the average number of years it will take for the securities in the 57 counties’
portfolios to mature.
Figure 4
Average Years to Maturity
for 57 County Portfolios
2.4 years or less
47 counties
Maturities in comparable
(82 percent)
investment funds average
2.4 years and Standard &
Greater than 3 years
Poor’s recommends that
4 counties
average maturities of (7 percent)
county and state portfolios
be limited to no more than
one year. 2.41 to 3.0 years
6 counties
(11 percent)
The average
Data based on survey responses by the various counties.
maturities reported by
62
the 543 counties managing their own investment portfolios during calendar year 1994 ranged
from 30 days to 4 years. The average maturity for a portfolio is the average length of time a
portfolio’s securities need to be held before they mature. According to our investment experts,
funds with comparable investment objectives to those of county treasurers’ portfolios maintain
portfolios where the average maturity for securities held is 2.4 years. Furthermore, Standard &
Poor’s recommends that the average maturity of county and state investment pools not exceed
one year. Figure 5 shows the average portfolio maturities during 1994 of more than 2.4 years
for counties that we reviewed.
Figure 5
Counties We Reviewed With Average Portfolio
Maturities of Over 2.4 Years During 1994
4.00
Colusa
2.55
Monterey
3.20
3.10
Placer
2.60
3.13 San Diego
3.60
2.74
Solano
2.90
0.0 2.0 4.0 6.0 8.0
Our review found that substantial investments in long-term securities significantly increased the
risk for two of the portfolios. For example, the Colusa County board of supervisors approved
the treasurer’s request to purchase investments with maturities longer than the five years
provided by law so as to increase the yield on investments, resulting in Colusa County investing
in high concentrations of securities with long maturities. This
strategy limited the county treasurer’s ability to restructure the
portfolio to provide additional cash without potentially realizing
a significant loss as of March 31, 1995.
The average maturity of the Colusa County portfolio was,
according to our investment experts, 27.9 years as of March
3 Three counties placed their funds in the Local Agency Investment Fund where investments can be
withdrawn on demand.
63
seitnuoC
28.40
Average Maturities as reported by the
counties for 1994.
Average maturities are based on
calculations from our investment experts
as of September 30,1994.
Years
Holding large
concentra-tions of
long-term securities is
more volatile than holding
short-term investments in
an environment of
changing interest rates.
1995, more than 11 times the average maturity appropriate for a portfolio with Colusa County’s
stated investment policy. According to our investment experts, the market value of the portfolio
has decreased to $19.4 million as of March 31, 1995, compared to the book value of the portfolio
of $24.3 million. Because of the long maturities and resulting higher duration of its securities,
the Colusa County portfolio is more volatile in an environment of changing interest rates than
portfolios with shorter maturities. Holding large concentrations of securities with long
maturities was a major cause of the portfolio’s sensitivity to interest rate changes and its decline
in market value as interest rates rose.
In another example, San Diego’s average maturity reached a high of 3.6 years during 1994. In
San Diego’s case, the rising interest rate market of late 1994 made the long maturity of the
county’s portfolio problematic. The market value of many of the county’s investments declined
below their original purchase price. To increase liquidity, the county sold securities before their
maturity and suffered a $7.9 million net loss to the portfolio.
High-Risk Investment Strategies Can
Result in Losses or Other Problems
The purchase of excessive concentrations of thinly traded structured securities, extensive
leveraging, and lengthy investment maturities is inconsistent with the basic investment principles
of safety, liquidity, and yield, and exposes the portfolios of local governments to an
inappropriate degree of risk. When an investment portfolio is managed in such a manner, the
investments are extremely sensitive to interest rate fluctuations, may not be readily converted to
cash, and may ultimately lose value.
As shown in Table 3, our investment experts performed an in-depth analysis of the eight
counties’ portfolios as of March 31, 1995. For each portfolio, they determined types of
structured investments, extent of leverage, and sensitivity to interest rate fluctuations (duration).
As previously discussed, duration is a quantitative measure of a portfolio’s sensitivity to a one
percent interest rate change. The effective duration combines the impact on the portfolio of
structured notes and the length of maturity, adjusted for leverage. According to our investment
experts, the duration for a comparable short-term, high-quality portfolio is 1.8. Our investment
experts computed an effective duration for all eight counties in our sample as of March 31, 1995,
and found that for six of the eight, the effective duration exceeded 1.8. They found that Colusa
County’s effective duration was 24.2. This duration means that, for every one percent increase
in interest rates, the market value of the base portfolio declines by 24.2 percent, or $4.2
million on a $17.2 million base. In Placer County’s case, the effective duration was 4.1,
which translates into a $12.1 million decline in the market value of the $296 million base
portfolio if interest rates increase one percent.
Table 3
Effective Durations for the
Investment Portfolios of Eight Counties
64
A B C D E F G
Nonstructured Structured
Nonstructured Fixed Rate Notes as a Effective
Fixed Notes as a Structured Percentage Total Duration
Rate Note Percentage Note of Total Portfolio Leverage of Base
County Date Duration of Total Duration Portfolio Duration Factor Portfolio
Portfolio
Colusa 3/31/95 2.7 30% 30.3 70% 22.0 1.10 24.2
Monterey 3/31/95 1.4 38 3.0 62 2.4 1.63 3.9
Placer 3/31/95 1.2 28 4.2 72 3.4 1.20 4.1
Sacramento 3/31/95 1.7 95 0.4 5 1.6 1.00 1.6
San Bernardino 3/31/95 1.7 67 0.7 33 1.4 1.26 1.7
San Diego 3/31/95 2.2 62 3.5 38 2.7 1.00 2.7
Solano 3/31/95 0.9 53 4.7 47 2.7 1.32 3.6
Sonoma 3/31/95 1.2 32 2.3 68 1.9 1.09 2.1
Note: E = (A x B) + (C x D)
F = Gross Portfolio Base Portfolio
G = E x F
Many of the Investments
Are Not Liquid
Purchasing structured securities and other investments that are
not actively traded also places a portfolio in a more risky
position because these securities often cannot be easily
Purchasing investments converted to cash to meet the needs of pool participants without
that are not actively traded incurring significant losses. To enhance liquidity, investments
places the portfolio in a
should be limited to securities that are actively traded in a
risky position because
secondary market. A secondary market is one in which there
thinly traded securities are
are sufficient buyers and sellers so that a particular asset can be
more difficult to convert to
readily converted to cash. In addition, more liquid investments
cash.
typically have a smaller spread between the price offered by the
buyer and the price asked by the seller. Liquidity levels over
time must enable the funds to meet forecasted cash
requirements, as well as provide immediate access to cash
needed for unexpected disbursements.
Our investment experts estimated the market value of each portfolio using a securities
information service for actively traded securities and a pricing model for thinly traded securities.
They found that seven of the eight counties held significant amounts of securities that were
thinly traded (not actively traded on the secondary market with a maturity greater than one year),
and, therefore, may not be readily convertible to cash.
As shown in Figure 6, 85 percent of Colusa County’s portfolio as of March 31, 1995, was
comprised of thinly traded securities. Furthermore, both Placer and San Diego Counties’
portfolios contained more than 60 percent of securities considered thinly traded.
Lack of liquidity can be a problem when the cash needs of participants or external demands such
as collateral calls exceed the immediate ability of the portfolio to convert its securities to cash.
For example, in Orange County, the demands of the investment brokers for additional collateral
65
exceeded the portfolio’s ability to meet those needs. As a result, the county was forced to file
bankruptcy.
Cash demands have been experienced by other counties to varying
degrees, forcing them to experience losses to meet those needs.
Unexpected cash
demands have forced Fortunately, none to date has been as dramatic as in Orange
some counties to County. However, in San Mateo County, the withdrawal of
experience losses to meet voluntary participants from the pool caused the county to
participants’ needs. prematurely sell securities for approximately $6 million less than
cost to meet their demands, resulting in reduced earnings to the
pool for two quarters. Figure 6
Thinly Traded Notes Maturing in
More Than One Year
for County Investment Portfolios
March 31, 1995
Moreover, as of
Sonoma 50%
September 30, 1994,
only 21.6 percent of the Solano 44%
market value of San Diego 65%
San Diego County’s
San Bernardino 20%
portfolio would have
Sacramento 5%
matured in less than one
Placer 70%
year. The county
imposed a restricted set Monterey 44%
of withdrawal rules in Colusa 85%
December 1994 when
0% 10% 20% 30% 40% 50% 60% 70% 80% 90%
voluntary participants,
Percentage of Total Portfolio
who represented
Note: Percentages based on par value totals.
38 percent of deposits,
expressed concern about
the viability of the pool and when they would be able to withdraw their funds. Thereafter, the
county implemented a plan, with the concurrence of its oversight committee, for a phased
withdrawal of voluntary participants through June 1998, thus averting the premature liquidation
of assets.
66
Portfolios Have Experienced
Unrealized Losses
As Figure 7 shows, all eight counties we reviewed have experienced unrealized losses ranging
from $4.8 million to $123 million as of March 31, 1995. Unrealized losses are the difference
between the book value and market value of a security and do not become realized unless the
security is sold.
The eight treasurers from these counties stated that they had sufficient liquidity in their portfolios
to preclude the sale of securities prior to maturity and, therefore, none believe that they will incur
any actual losses.
Figure 7
Counties Experiencing Unrealized Losses
March 31, 1995
San Diego $3,0 $ 0 3 2 ,125 (-3.94%)
San Bernardino $2 $ , 2 0 , 4 1 4 02 (-2.76%)
Sacramento $1,075 (-2.89%)
$1,107
Sonoma $ $ 6 6 6 9 9 0 (-3.04%)
Monterey $524 (-5.07%)
$552
Placer $358 (-5.29%) Market Value
$378
Book Value
Solano $342 (-4.47%)
$358
$19
Colusa $24 (-20.83%)
$0 $500 $1,000 $1,500 $2,000 $2,500 $3,000 $3,500
Portfolio Value (In Millions)
67
Blank page inserted for reproduction purposes only.
68
Chapter 2
Current Regulations and Guidelines Fail To
Address Investments Made By Agents
Chapter Summary
Six of the 57 counties we surveyed responded that they use agents, many also acting as security
custodians, to execute securities lending or repurchase agreement transactions. We reviewed
the securities lending and repurchase activities of three of these counties. These agents make
investments on behalf of the counties by negotiating securities lending or repurchase agreements
and purchasing various other securities with the proceeds. When these agents’ transactions are
in effect, certain characteristics of ownership pass from the county.
Currently, there are no laws, regulations, guidelines, or standards requiring county treasurers to
record these transactions in their accounting records or to inform their pool participants or boards
of supervisors regarding securities lending or repurchase transactions conducted by agents.
Additionally, the California Government Code does not address securities lending or repurchase
transactions conducted by agents by either authorizing such agent activities or by setting
investment criteria relating to these activities. Therefore, they are considered “off-book”
investments.
Three counties we reviewed used agents to execute securities lending and repurchase
transactions. These counties’ investment pools bear all the market risk of investments
transacted by agents while the profits are shared with the agents for services that, in one case,
was not competitively bid. Finally, at two of the counties we reviewed, agents were allowed to
transact investment business with their affiliates.
As a result of investing in this manner, counties have no assurance that state investment controls
are being met, local investment participants and legislative bodies lack information about the
actual composition or risk in the investment portfolio, and potential conflicts of interest are
created when transactions are allowed with affiliates.
Agents Convert County Securities
to Cash or Other Investments
69
In 6 of the 57 counties in the State, the counties use an agent to execute some securities
transactions. These agents execute investment transactions after being delegated the authority to
do so by the counties. Three of the 8 counties we reviewed (Sacramento, San Diego, and
Sonoma) use agents to execute securities transactions. Each county’s agreement is a securities
lending and/or repurchase agreement. However, regardless of the title of the agreements entered
into by the three counties, we found that the economic substance of the transactions resulting
from the agreements was similar. Namely, the counties are using agents to convert county
securities to cash, which the agents reinvest, or to trade the securities for different ones and a
lending fee. We will refer to all of these transactions as
securities lending.
Regardless of the title, Securities lending, when an agent is used, involves three parties:
under these agreements, the county, an agent for the county such as a bank or other
counties are using agents financial institution, and a counterparty (borrower). The agent
to convert securities to loans these securities to a borrower on behalf of the county. The
cash for reinvestment. borrower secures the loan with either cash or securities equal to
100 percent of the market value of the securities borrowed plus a
margin of 2 percent. As a result, the borrower has use of the
borrowed securities. When securities are pledged as collateral
for a loan, this is a “true” securities lending transaction. In this
instance, the borrower pays a negotiated fee based on the value of the security borrowed and the
loan term. Thus, the return on the loan is a lending fee. The lending fee is split between the
county and agent in a predetermined ratio.
When the collateral pledged by the borrower is cash rather than another security, the transaction
has a second phase. This second phase is economically equivalent to a reverse repurchase
transaction from the county’s perspective since the county’s agent reinvests the cash by
purchasing another security. Also, a specified rate of interest is paid to the borrower of the
security for the cash collateral provided to the county; the interest paid is the “rebate.” So, to
offset the rebate, the cash is reinvested and the difference between the interest payments on the
purchased security and the rebate is the return on the loan transaction. Any return generated is
split on a predetermined ratio between the agent and the county.
Agents attempt to generate a return through one of three methods. First, they can lend a security
with a short maturity and invest in a security with a longer maturity and higher interest rate.
Second, they can lend a security with a high credit rating and invest in a security with a lower
credit rating and, typically, a higher interest rate. Third, agents can lend securities that are in
high demand and either receive a high loan premium from the borrower or pay the borrower a
lower rebate rate.
There are some common features to a securities lending transaction regardless of the type of
collateral pledged. The agent assesses the market value of the loaned security daily and
requires additional collateral if the collateral’s market value falls below the market value of the
loaned security. The county generally bears the risk of any loss of principal from investing cash
collateral. The opposite is also true—collateral is returned to the borrower if its market value
increases above the market value of the loaned security.
70
Typically, a lending agreement can call for the transaction to be terminated at either a
predetermined date or at the county’s discretion. However, while the transaction is in effect,
certain characteristics of ownership pass from the county to the borrower. While counties enter
into lending transactions to earn additional income, borrowers (for example, broker-dealers and
financial institutions) enter into lending agreements for several reasons. For example, the
borrower can use the security for delivery to cover a short sale
(i.e., the broker sold a security without owning it), to meet a
contractual obligation to purchase a certain security for a specific
In a lending agreement, buyer, or to pledge the security as collateral for another
certain characteristics of transaction. While the lending agreement is in effect, the
security ownership pass counties do not maintain these characteristics of ownership over
from the county to the
the securities they have loaned. However, since they have the
borrower.
contractual right to receive identical, but not the same, securities
at the end of their loans, the counties continue to benefit from
any increase in the securities’ market value or bear any loss from
a decrease in market value. Also, although counties no longer
receive interest payments directly from debt issuers, borrowers who hold the securities are
contractually obligated to pass any interest payments they receive to the counties during the loan
term.
As noted above, counties commonly enter into securities lending transactions to earn additional
income on their portfolios. However, as with most types of investments, there are risks
involved. For example, there is credit risk because the borrower may go out of business and be
unable to return the security to the county. In addition, there is market risk. The county has an
obligation to return cash collateral in the amount received. If the market value of the
reinvestment falls below the amount of the cash collateral that was pledged by the borrower, the
county is responsible for making up the difference to meet its obligation to return cash collateral
in the amount received.
Agent Transactions Are “Off-Book”
During our review of the three counties noted previously, we found that the transactions
conducted by agents were all “off-book.” At each of the three counties we reviewed that use
agents, the county treasurers all receive monthly reports from their agents summarizing the
counties’ lending or repurchase positions. Sacramento, Sonoma, and San Diego counties
receive similar monthly information providing daily transaction and lending and reinvesting
activity. The securities lending reports include data about the securities on loan, the borrowers,
loan amounts, types of collateral pledged, and a description of the reinvestments made on the
county’s behalf including the rates of return. However, none of these transactions are ever
recorded in the financial records of the counties, and are thus “off-book.” Partially because
these transactions are “off-book,” they are not treated the same as other county transactions. For
example, we found that the securities lending activities are not
disclosed in the periodic reports to boards of supervisors or pool
participants. Also, although the counties recognize the
Although counties receive
importance of the quality of the reinvestments made with the
information from agents
collateral received for a securities loan, they do not incorporate
on security lending
transactions, they are not the reinvestment in compliance reviews of their investment
required to disclose this portfolios.
information in periodic
reports or in financial
records.
71
Various factors contribute to securities lending transactions being “off-book”. For example,
there are no current requirements for county treasurers to inform pool participants of securities
lending transactions. Also, since the California Government Code does not specifically
authorize securities lending, it also does not specifically state how these transactions relate to
requirements stated in the code. In addition, there is no currently applicable standard for
reporting on securities lending transactions issued by the Governmental Accounting Standards
Board related to external financial reports.
Boards of Supervisors and Pool
Participants are not Periodically Informed
of Securities Lending Activity
County treasurers are not currently required to periodically inform their board of supervisors or
pool participants of the investment activities of the investment pool or of their lending agent.
However, many county treasurers do report to their board or pool participants regarding
investment activities and yields. Although the Sacramento, San Diego, and Sonoma county
treasurers prepared these periodic reports, the reports we reviewed did not discuss securities
lending. For example, the reports did not disclose that some securities in the portfolio had been
loaned, that other securities with different maturities or credit qualities may be held as collateral
backing pool deposits rather than the loaned securities, or clearly identify the earnings resulting
from lending transactions. The extent of securities lending at the three counties was significant;
as shown in Figure 8, at September 30, 1994, the three counties had loaned from 23 percent to
59 percent of their respective portfolios. Therefore, the periodic reports do not contain complete
data on the investment activities of the counties’ pools.
Agent Transactions Could Avoid
Statutory Limitations and are
Excluded From Compliance Reviews
In our March 1995 report entitled, Orange County: Treasurer’s Investment Strategy Was
Excessively Risky and Violated the Public Trust, the Bureau of
State Audits recommended a statutory 20 percent limitation on
reverse repurchase agreements, and limiting to 5 percent of this
On the surface, a portfolio
amount, investment for yield enhancement. Legislation is also
could appear to comply
pending that will restrict reverse repurchase agreements and
with statutory provisions;
impose other constraints on investment activities. However,
however, with off-book
transactions, any such counties could inappropriately avoid future investment
limitations could be limitations when employing agents to conduct off-book
circumvented. investments. On the surface, under current disclosure practices,
a portfolio could appear to comply with statutory provisions;
however, because agent transactions are not reflected in
portfolios, any such limitations could easily be circumvented
without detection. Consequently, under these circumstances, statutory limitations are
meaningless.
72
Figure 8
Size of County Portfolios and Amounts
and Percentages Loaned by Agents
September 30, 1994
$4,000,000
$3,690,695
$3,500,000
$3,000,000
$2,500,000
$2,000,000
$1,500,000 $1,094,585
$1,073,783 30%
$1,000,000 $633,532
59% $613,506
$500,000 $142,365
23%
$0
Sacramento San Diego Sonoma
The California Government Code, Sections 53600 et. seq., directs the investment activities of
county treasurers. Intending to minimize portfolio risk, the code establishes the allowed types
of investments, the holding concentrations, the maximum maturities, and the credit requirements
of securities issuers. Since the code does not specifically authorize securities lending, it also
does not specifically state how securities lending transactions relate to requirements stated in the
code. However, it does authorize counties to engage in repurchase and reverse repurchase
transactions. At the eight counties we visited, we found that they actively review their portfolio
holdings for compliance with statutory regulations. Further, the three counties we reviewed that
use securities lending agreements recognize the importance of the quality of the reinvestment
and incorporated limitations similar to those in the code in their securities lending agreements.
Each agreement specified the types of investment instruments allowed for reinvestment and
imposed credit qualifications for the securities issuers. These specifications are important; if the
borrower does not return the borrowed security at the end of the loan, the reinvestment may
replace the loaned security. However, none of the compliance reviews conducted at the three
counties we reviewed—Sacramento, San Diego, and Sonoma—included security holdings
resulting from the investment activities of the counties’ agents. As a result, if the majority of
reinvestments holdings are in medium-term commercial paper and the county also invested in a
significant amount of commercial paper, the county could be unknowingly backing pool deposits
with a larger percentage of commercial paper holdings than is allowed by law.
73
)sdnasuohT
nI(
eziS
oiloftroP
Portfolio Book Value at 9/30/94 Book Value on Loan
Securities Lending Transactions
Are Not Required To Be Fully
Disclosed in Financial Statements
The Government Accounting Standards Board (GASB) has set standards for governmental
financial reporting and disclosure. Currently, GASB Statement No. 3, entitled “Deposits with
Financial Institutions, Investments (including Repurchase Agreements), and Reverse Repurchase
Agreements,” requires footnote disclosure in financial statements related to investments.
However, neither governmental accounting standards nor State law or regulations address
accounting and financial reporting for securities transactions
conducted by an agent, and, therefore, no required financial
statement disclosure for agents’ transactions is required. All
One county had up to
transactions made by agents are “off the books”. While the
58 percent of its portfolio
security is loaned, a county’s agent is holding a security
on loan throughout 1994;
unrelated to the one placed in its custody. However, the
neither its internal financial
county’s balance sheet continues to reflect the original security
records nor its reports
were required to disclose placed in custody, not the one currently held by the agent. For
this example, personnel from Sacramento County’s treasury stated
activity. that up to 58 percent of the county’s pooled investment portfolio
was on loan throughout 1994; however, neither the county’s
internal financial reports nor its Comprehensive Annual
Financial Report reflect this activity. In the Sonoma County’s
Comprehensive Annual Financial Report, the county discloses the amount of securities on loan at
year end and a description of the securities lending program. However, the county is not
required to and it does not disclose the amount or the nature of the securities the agent purchased
with the proceeds from the securities that were loaned.
GASB Statement No. 3 poses an additional consideration related to custodial credit risk
classification. This statement classifies securities in three custodial credit risk classifications as
follows: (1) securities insured or registered or held by the entity or its agent in the entity’s
name; (2) uninsured and unregistered with securities held by the counterparty’s trust department
or agent in the entity’s name; and (3) uninsured and unregistered, with securities held by the
counterparty, or by its trust department or agent but not in the entity’s name. Except for
Sonoma County, which did not classify securities that were loaned, the footnote classification as
of June 30, 1994, in the most recently issued annual financial reports for the counties we
reviewed, reflected the custodial credit risk classification of the original security investment.
The risk classifications were generally 1, the safest classification available. However, as in the
case of San Diego County, with an average of 34 percent of its total portfolio on loan, the
classification disclosed in the financial statements may not represent the true nature of the
custodial risk at the report date. Although San Diego County’s financial statement disclosures
may comply with currently applicable requirements, the true custodial credit risk classification of
the county’s investments are likely to be category 2 or 3 since the securities may not be in the
county’s name and the custodian may be acting as the agent for, or be affiliated with, the
counterparty.
74
Counties’ Investment Pools Bear Entire
Market Risk, Agents Share in Profit
We found that at all three of the counties we reviewed that deal with agents, the counties’
investment pools, not the agents, bear all the losses due to market risk. For example,
agreements for Sonoma and San Diego counties specifically state that “All Approved
Investments are for the account and risk of Customer. To the extent of any deficiency in the
amount of Collateral available for return to (a borrower) due to
losses on Approved Investments or otherwise, Customer shall
deliver to Bank . . . funds in an amount equal to such
In Lending Agreements,
deficiency. Bank may, but is not obligated to, pay any such
counties are solely liable .
deficiency . . . and all such payments shall constitute loans . . . to
. . the county’s risk
Customer . . . .” This passage means that the county, as a party
exposure is called market
to the agreement, must return the total collateral held by its agent
risk.
to the borrower at the conclusion of the agreement. If the value
of the collateral invested by the agent falls below the amount
owed to the borrower, the difference must be contributed by the
county to meet the obligation. The county’s risk exposure is called market risk. The counties
are solely liable to provide the funds needed to meet their collateral obligations to borrowers.
On the other hand, we found that the agreements allow the agents to share in all the profits
resulting from lending fees or the difference between the interest payments received on the
reinvestments and the rebates paid on the loaned securities. In the three counties’ agreements
we reviewed, the agents earn between 30 and 50 percent of the net profits from the activities of
investing the counties’ securities. As a case in point, Sacramento County’s agent has full
lending access to the county’s portfolio and receives 30 percent of the proceeds of the lending
transactions. In 1994, Sacramento County earned approximately $878,000 from its securities
lending activities; we calculated the agent’s share of the profits to be approximately $375,000,
excluding any fees and commissions the agent may have earned.
All Counties Are Not Competitively
Bidding for Agent Services
Each of the counties we reviewed use their custodian as their agent. A reason counties cite for
using their custodians as agents is to eliminate or reduce
custodial fees. Also, because loss due to market risk is borne
solely by the county, it is more in a custodian’s interest to act as
It is more in a custodian’s
both custodian and agent. To illustrate, Sacramento County’s
interest to act as both
investment personnel, who manage a portfolio of roughly $1.1
custodian and agent.
billion, estimate custodial fees would cost the county $140,000
annually. By allowing its custodian to also act as its agent,
Sacramento County was able to negotiate an annual maximum
custodial fee of $40,000.
When describing their securities lending activities to us, the three counties did not state that they
competitively bid for the services of a securities lending agent separate from their custodial
services. However, Sacramento and San Diego counties did competitively bid for combined
securities lending and custodial services. An agent engaging in securities lending acts very
75
much like an investment manager. The agent is responsible for negotiating loan terms,
evaluating the credit quality of the borrowers, and developing and executing a strategy for
investing collateral. The fee paid to the agent should be based on the agent’s proven ability to
perform in these areas. As a result, by not competitively bidding for the services of securities
lending agents, one of the counties cannot demonstrate that its agent’s share of the profits is not
greater than it needs to be. For example, Sacramento and San Diego counties, who did
competitively bid for their securities lending services, pay their agents 30 and 40 percent,
respectively, of the net profits from lending, yet Sonoma, who did not competitively bid its
securities lending services, pays its agent 50 percent.
Additional Risks to Counties
Are Not Explicitly Covered
In Contracts With Agents
Counties add an additional aspect of risk when using agents because the counties effectively
relinquish control over their portfolios. In the lending agreement with Sacramento County and
its agent, the agent has access to 100 percent of the county’s portfolio for lending purposes.
Sonoma County’s agreement does not address this topic directly but states “. . . (c)ustomer
hereby appoints Bank as its agent to lend Securities in the Custody Account . . . .” This
statement implies total access. When agents have the ability to invest on the counties’ behalf
and have total access to the counties’ portfolios, agents may hinder the counties’ abilities to sell
specific securities.
For example, if a county treasurer decides to sell certain securities to take advantage of favorable
market conditions, the securities may be out on loan and not immediately retrievable. The
counties we reviewed that engaged in securities lending through agents informed us that
securities would be and have been returned upon request. Sonoma County’s agreement states
that securities will be returned “as soon as practicable” after receipt by the bank of written
instructions from the county. Sacramento County’s lending agreement provides for the return of
securities on the same day, or the next, depending upon the time of day the request is made.
The three agreements we reviewed did provide the counties protection in a number of ways;
however, the agreements did not specifically protect the counties in the event they were unable to
complete the sale of a security because their agents were unable to deliver a security on the day it
was needed.
Agents Dealing With Affiliates
May Create Conflict of Interest
The use of an agent who invests with affiliates may create an
Agents for two counties
actual or potential conflict of interest. In the agreements for
have the sole discretion to
San Diego and Sonoma counties that we reviewed, the agent was
deal with affiliates and to
allowed to invest with affiliates; that is, in the course of investing,
pay them the commissions
the borrowers can be affiliated with the county’s agent.
and fees the agent deems
appropriate. Moreover, the agreements authorize the agent to pay customary
fees and commissions on these activities. The authorized dealing
with affiliates raises ethical and prudent investment issues.
76
While we did not audit the activities of the agents and, therefore, have no information suggesting
an actual conflict of interest, dealing with affiliates, nevertheless, gives agents the opportunity to
earn income on all sides of the investment transaction. Therefore, an unscrupulous agent could
be encouraged to “churn” the county’s portfolio to generate revenue primarily from commissions
and fees rather than from investment proceeds that are shared with the county or from
investments that increase county yield while minimizing county risk. Furthermore, the ability to
deal with affiliates provides the agent the opportunity to steer business to an affiliate that
otherwise might not have been competitive enough to win the business on its own. Moreover,
the counties’ agreements state that dealings with affiliates are at the sole discretion of the agent
and that commissions, fees, or amounts paid for securities are as the agent deems appropriate.
However, the agreements do state that in these transactions, affiliates will be treated no
differently than if they were unaffiliated parties. While this provision offers some protection,
it still would not preclude the churning of accounts or steering of business to an affiliate.
77
Blank page inserted for reproduction purposes only.
78
Chapter 3
Recommendations
D
uring our audit of the investment practices of county treasurers, we encountered some of
the same risky investment strategies found during our earlier audit of Orange County,
namely, overinvesting in structured notes, excessive leveraging of portfolios, and
investing in large quantities of long-term securities. Therefore, it seems appropriate to reiterate
for legislative consideration the same recommendations made to Orange County.
The Legislature should amend the California Government Code related to local
government investment practices to do the following:
Require written investment policies for all local entities investing public funds that are
approved and adopted by their local governing body. These policies should ensure that
safety and liquidity are paramount to yield;
Limit the use of reverse repurchase agreements to no more than 20 percent of the total
portfolio, primarily to meet immediate or unexpected cash flow requirements, and not for
reinvestment. In no case should the use of reverse repurchase agreements or other types of
borrowing for yield enhancement or risk arbitrage exceed more than 5 percent of the
portfolio. Further, multiple levels of borrowing should be prohibited;
Establish and define a prudent person rule for the local investment officer. The prudent
person rule should detail the fiduciary responsibilities vested in the investment officer and
establish an expected level of expertise;
Limit the use of derivatives or other structured investment instruments and prohibit those that
put principal at risk. None of these instruments are to be purchased with borrowed or
leveraged funds. Further, the derivatives or structured investments purchased should be
openly traded in the secondary market on a recognized exchange. Any investments in these
instruments should be limited to no more than 5 percent of the portfolio;
Require that investment officers consider purchasing securities receiving a favorable
volatility rating from a nationally recognized credit rating agency, whenever possible.
These ratings provide important information about an investment by assessing risk over a
wide range of conditions, including the effect of interest rates, prepayments, credit, spread,
liquidity, and currency fluctuations;
Impose limitations on the average length of maturity for local government investment
portfolios to meet cash flow requirements and liabilities;
Require competitive bidding or pricing for all investments purchased and mandate that the
investment officer maintain a history of the competitive bidding and selection;
79
Mandate investment reports at least quarterly to the governing body and investment pool
participants that include detail of the inventory and transactions during the period, weighted
average maturity, current market value, duration or other similar interest rate sensitivity
analyses, and yield calculation of the portfolio; and
Prohibit the issuance of taxable or nontaxable debt for the purpose of investing the proceeds
in an investment pool or purchasing an investment security for speculation or risk arbitrage.
The Legislature should also amend the California Government Code related to local
government investment practices to do the following:
Limit the average maturities of securities held in county portfolios for short-term, operating,
and surplus cash investments to 2.4 years or less;
Competitively bid separately for lending agent and custodial services, and ensure that all
county investment activity by agents or the county are properly recorded and disclosed to
interested parties;
Require counties to maintain sufficient liquidity to match the average amount deposited by
voluntary participants of the investment pool and to meet the cash flow needs of their
involuntary members; and
In limiting the use of derivatives or other structured investments to no more than 5 percent of
the portfolio, the prohibition should be on those structured notes that increase risk to the
portfolio as measured by a derivative analysis.
We recommend that the county treasurers employing high-risk strategies do the following:
Avoid the use of these risky strategies in the future;
Prudently divest themselves of structured notes that add risk to their respective portfolios;
Limit future purchases of structured notes that increase risk to the portfolio (as measured by a
duration analysis) to no more than 5 percent of all investments;
Confine the use of leverage in their respective portfolios to 20 percent or less; and
Prudently restructure their respective portfolios to reach an average maturity of no more than
2.4 years.
Further, those county treasurers who participate in a security lending program should do
the following:
Include agent transactions in statutory limitations and compliance reviews;
Report on the security lending program in the counties’ annual financial reports and in
reports to boards of supervisors and pool participants;
80
Competitively bid separately for lending agent and custodian services; and
Preclude agents from dealing with their affiliates.
In addition, to comply with State law, the specific treasurers listed below should do the
following:
The Colusa County treasurer should follow the law in allocating all interest earnings to pool
participants and not retain a portion of interest to offset declines in the market values of
securities in the portfolio; and
The Sacramento treasurer should allocate interest earnings to all pool participants on a
quarterly basis and recoup the overallocation made to one pool participant and reallocate the
amount to all participants in accordance with the law.
81
We conducted this review under the authority vested in the state auditor by Section 8543 et seq.,
of the California Government Code and according to generally accepted governmental auditing
standards. We limited our review to those areas specified in the audit scope of this report.
Respectfully submitted,
KURT R. SJOBERG
State Auditor
Date: June 13, 1995
Staff: Philip Jelicich, Audit Principal, CPA
Doug Cordiner
Arthur Monroe, CPA
Colleen Schroeder, CPA
Jaana Brown
Sharon Smagala
Michael Tilden, CPA
Dave Frizzell
Christiana Mbome
Phyllis Miller, CPA
Appendix A
Interest Allocations That Deviate
From Government Code Requirements
T
he treasury offices of Colusa and Sacramento counties did not allocate interest to pool
participants in compliance with the California Government Code, Section 53684(b).
Specifically, the code requires county treasurers to apportion quarterly any interest from
the investment of funds in an amount proportionate to the average daily balance of the amounts
deposited. Our review disclosed the following:
Colusa
Colusa County apportions interest to pool participants quarterly. However, the treasurer is not
allocating the total interest earnings received. Instead, the treasurer considers some earnings as
a return of principal and adjusts the book value of the securities to more closely reflect current
market value, thus attempting to minimize any loss in the portfolio. This activity reduces
the total interest to be apportioned to pool participants. For the quarter ended
September 30, 1994, the treasurer reclassified approximately $46,000 of the interest earned
rather than apportioning it to the 47 pool participants.
47
Sacramento
Sacramento County’s policy is to apportion interest to pool participants annually rather than
quarterly as required. Specifically, all agencies and districts, except the Coroner’s Office and
the agencies included in the Local Agency Investment Trust (LAIT), receive interest earnings
annually at the end of the fiscal year. The Coroner’s Office and LAIT receive interest quarterly
because their agreements with the county require the quarterly apportionment of interest in
compliance with the California Government Code.
In addition, the auditor-controller’s office overallocated $16,720 of the total $60,930,277 in
distributed interest earnings to the Cable Commission, to the detriment of the remaining pool
participants. Specifically, a portion of the Cable Commission’s monies were invested and
earned interest in a separate portfolio. However, the average daily balance for the separate
investment was inappropriately included in the average daily balance for the treasurer’s
portfolio. As a result, the average daily balance and the interest apportioned to the Cable
Commission was overstated.
Appendix B
Selected Comparable Financial Data
of County Investment Portfolios for 1994
Average Average
Percentage Percentag
Book Value Market Value Annual Yield Average of Portfolio e
of Net of Net Based on Portfolio in of
Investment Investment Quarterly Maturity Structured Portfolio
County June 30, 1994 June 30, 1994 Apportionment in Years Securities Leveraged
s (Leverage
Ratio)
1 Alameda $ 1,281,983,000 $ 1,281,983,000 B 4.03% 1.97 6.00% 0.50%
2 Alpine 7,609,367 7,609,367 4.42 0.00
D
3 Amador 33,276,700 32,874,685 3.08 0.52
E
4 Butte 120,282,401 120,110,254 5.03 1.33
5 Calaveras 44,293,194 44,114,210 4.15 0.85 2.28
F
6 Colusa 18,183,913 A 16,692,000 A 7.32 4.00 30.00 20.00
7 Contra Costa 863,489,000 862,601,000 4.59 0.70 15.53 9.99
8 Del Norte 16,701,089 16,701,089 B 5.05 0.62
9 El Dorado 127,460,476 127,537,126 3.17 0.37
10 Fresno 949,853,299 932,933,452 4.63 0.71
11 Glenn 19,002,239 19,002,239 B 4.21 1.70
12 Humboldt 108,448,227 106,670,452 6.17 2.05
13 Imperial 128,818,319 128,736,302 5.01 2.87 20.00 H
14 Inyo 27,263,956 27,199,183 3.98 0.25
15 Kern 836,379,000 833,108,000 4.41 1.03 6.50 4.00
16 Kings 92,667,921 91,118,434 4.48 1.13 7.77 8.50
17 Lake 57,258,578 57,124,023 3.50 0.31 6.00
18 Lassen 29,436,794 28,560,714 4.23 3.54
48
F
19 Los Angeles 6,917,014,000 6,867,705,000 4.33 0.78 8.28 3.00
20 Madera 96,000,000 A 96,000,000 A 3.62 1.72 4.00
21 Marin 206,408,248 206,259,248 4.05 0.43
22 Mariposa 15,124,475 14,650,088 4.34 2.88
23 Mendocino 37,217,822 37,417,422 3.99 0.45 6.00
24 Merced 184,408,808 A 184,408,808 A 3.42 0.47 7.00 19.00
E
25 Modoc 12,490,514 12,490,514 3.16 0.00
D
26 Mono 16,124,055 16,124,055 B 4.07 0.00
D
27 Monterey 327,340,078 A 317,545,613 A 5.37 2.55 22.00 43.80
28 Napa 130,180,096 130,302,827 4.37 1.83 9.80
29 Nevada 77,570,955 C 75,288,647 C 5.85 0.55
30 Placer 342,350,326 320,648,964 6.77 3.10 52.00 37.26
31 Plumas 35,975,169 35,076,230 4.34 1.53
32 Riverside 1,079,220,000 1,046,282,000 3.72 2.62 1.20 2.90
33 Sacramento 1,720,212,000 1,720,212,000 B 5.02 2.48 1.00
34 San Benito 14,884,071 A 14,884,071 A 4.01 .27 - .36 10.00
35 San Bernardino 1,680,436,000 1,638,434,000 5.05 2.13 30.33 G 28.60
36 San Diego 3,289,275,000 3,289,275,000 5.18 3.13 34.50 1.45
37 San Francisco 2,211,797,845 2,207,506,010 5.55 2.09
38 San Joaquin 311,413,000 311,321,000 3.92 0.08
39 San Luis Obispo 253,681,265 253,221,123 3.99 0.38
40 San Mateo 630,967,380 612,215,400 3.07 1.75 12.00
41 Santa Barbara 378,756,238 379,790,208 4.68 0.99
42 Santa Clara 1,579,328,000 1,570,085,000 4.33 0.92 3.77
49
Appendix B Table
continued
Average Average
Percenta Percenta
Book Value Market Value Annual Yield Averag ge ge
of Net of Net Based on e of of
Investment Investment Quarterly Portfoli Portfolio Portfolio
County June 30, 1994 June 30, 1994 Apportionme o in Leverage
nts Maturit Structure d
y d (Leverag
in Securities e
Years Ratio)
4 Santa Cruz 217,417,047 218,454,686 4.18 0.51 0.65
3
4 Shasta 116,362,415 116,153,302 3.99 0.52
4
4 Sierra 6,690,317 6,647,941 5.72 1.75
5
4 Siskiyou 34,989,887 34,103,934 4.24 1.78
6
4 Solano 230,544,291 221,902,156 5.85 2.74 32.30 40.00
7
4 Sonoma 819,755,719 815,106,175 5.46 2.33 61.00 G 13.50
8
4 Stanislaus 323,350,079 318,848,606 4.54 1.92
9
5 Sutter 56,506,171 56,084,391 4.47 1.69 11.70
0
5 Tehama 36,075,242 34,751,572 4.47 2.20 31.00 20.00
1
5 Trinity 21,823,266 21,689,100 4.08 0.28 8.50
2 F
5 Tulare 305,936,291 302,943,628 4.27 2.00 1.00 10.00
3
5 Tuolumne 35,491,556 35,462,368 4.08 0.26
4 E
5 Ventura 671,072,188 665,640,032 4.18 0.49
5
5 Yolo 110,781,506 109,944,710 3.17 0.55 10.00
6
5 Yuba 107,868,361 105,413,913 4.63 2.35 16.80 25.00
7
Total $29,405,247,15 $29,154,965,272
Average 4 4.47% 1.38 16.65% 14.61%
160
Legend:
A Audited financial statements were not yet available; thus, if an estimate was not provided, we used book value for
market value.
B Market values for the treasurer’s investment pool cannot be identified from the financial statements; thus, we used
book value for
market value.
C Book and market values as of December 31, 1994.
D Investment funds are kept in the State of California Local Agency Investment Fund (LAIF) where investments can
be withdrawn daily
or on demand.
E The portfolio maturity as of December 1994.
F The portfolio maturity as of June 30, 1994.
G As of September 30, 1994, based on our investment experts’ analysis because the county did not provide us an
average for the year.
H County originally reported 28 percent and later amended its response to 4.75 percent. We audited this figure and
found that the percentage of structured notes on June 30, 1994, and September 30, 1994, was approximately
22 percent and 19 percent, respectively.
161
Insert 1st page of Appendix C
162
Insert 2nd page of Appendix C
Appendix C
Counties’ Investment Pool Participants
by Type and Average Percentage of the Pool
Number of Number of
Special Special
Number Number of Districts Districts
of Average Community Average Number Average Required To Average Allowed To Average
School Percentage College Percentage of Percentage Use County Percentage Choose Percentage
County Districts of Pool Districts of Pool Cities of Pool Treasury of Pool Treasury of Pool
1 Alameda 18 50.00% A 3 N/A 0 N/A A 10 N/A 1 N/A A
A A
2 Alpine 1 35.00 0 0.00% 0 0.00% 3 2.00% 1 0.07%
3 Amador 1 9.00 0 0.00 0 0.00 8 6.00 9 4.00
4 Butte 16 40.00 1 1.00 5 0.00 13 8.00 18 2.00
5 Calaveras 3 23.58 0 0.00 0 0.00 17 3.44 10 4.42
6 Colusa 6 33.18 1 3.08 2 0.26 38 22.72 34 22.74
7 Contra Costa 18 18.80 1 2.50 0 0.00 3 0.02 11 16.20
8 Del Norte 1 31.00 0 0.00 0 0.00 4 6.00 11 11.00
9 El Dorado 15 29.52 3 0.82 2 0.79 105 9.51 14 2.58
10 Fresno 48 35.00 4 0.40 0 0.00 75 8.40 22 2.80
11 Glenn 10 32.90 0 0.00 0 0.00 7 2.50 26 17.80
12 Humboldt 34 41.30 2 5.06 7 1.75 2 0.06 18 3.91
13 Imperial 10 40.00 B 1 N/A 0 0.00 16 3.00 0 0.00
B
14 Inyo 7 12.00 1 1.00 1 1.00 6 1.00 0 0.00
15 Kern 46 37.37 3 1.42 0 0.00 20 2.72 10 1.17
16 Kings 14 48.80 0 0.00 0 0.00 28 6.04 20 9.47
17 Lake 7 17.00 0 0.00 0 0.00 35 15.00 0 0.00
18 Lassen 11 46.00 1 4.00 0 0.00 4 1.00 21 2.00
19 Los Angeles 79 33.00 13 1.00 30 2.00 142 4.20 35 23.00
20 Madera 10 10.00 2 0.00 2 0.00 3 1.00 5 1.00
21 Marin 19 25.00 1 2.00 0 0.00 71 7.00 20 15.00
22 Mariposa 1 11.25 0 0.00 0 0.00 0 0.00 0 0.00
23 Mendocino 16 31.92 1 2.04 0 0.00 12 0.86 0 0.00
24 Merced 22 17.28 1 1.00 0 0.00 22 3.00 52 3.00
25 Modoc 4 49.50 0 0.00 0 0.00 10 0.02 17 0.01
26 Mono 2 17.00 0 0.00 1 0.00 2 0.00 16 16.00
27 Monterey 27 49.40 B 2 N/A 0 0.00 20 3.70 22 6.20
B
28 Napa 8 24.30 1 2.12 0 0.00 17 21.97 1 2.39
29 Nevada 10 41.20 1 0.02 3 2.75 N/P N/P N/P N/P
30 Placer 19 39.00 1 3.20 2 2.10 8 0.93 16 0.53
31 Plumas 1 19.00 1 1.55 0 0.00 8 1.66 37 7.08
32 Riverside 23 27.90 4 1.10 0 0.00 74 2.40 94 12.63
33 Sacramento 17 18.20 1 0.40 2 0.06 13 17.30 53 4.50
34 San Benito 11 58.75 0 0.00 0 0.00 N/P N/P N/P N/P
35 San Bernardino 41 31.89 4 1.23 2 4.29 6 0.13 9 0.08
36 San Diego 43 23.30 5 2.50 15 7.60 49 1.63 22 14.80
Appendix C Table continued
Number of Number of
Special Special
Number Number of Districts Districts
of Average Community Average Number Average Required To Average Allowed To Average
School Percentage College Percentage of Percentage Use County Percentage Choose Percentage
163
County Districts of Pool Districts of Pool Cities of Pool Treasury of Pool Treasury of Pool
37 San Francisco 1 10.43 1 0.46 1 N/A 0 0.00 0 0.00
38 San Joaquin 26 34.00 2 2.00 7 0.20 124 N/P 99 N/P
39 San Luis Obispo 13 30.00 1 1.00 0 0.00 16 4.00 0 0.00
40 San Mateo 26 24.00 1 1.50 21 1.00 33 8.00 N/P N/P
41 Santa Barbara 24 15.30 2 1.51 0 0.00 45 12.15 5 2.85
42 Santa Clara 35 30.60 4 3.38 0 0.00 9 2.20 18 9.30
43 Santa Cruz 14 26.60 1 1.72 4 1.61 3 0.04 12 3.22
44 Shasta 27 52.00 B 2 N/A 3 0.00 11 5.00 C 22 N/A C
B
45 Sierra 1 23.60 0 0.00 0 0.00 3 1.20 4 0.60
46 Siskiyou 29 2.73 1 0.15 0 0.00 27 2.45 1 0.05
47 Solano 6 23.00 1 1.00 0 0.00 9 2.34 12 0.01
48 Sonoma 42 26.97 1 3.46 8 9.14 4 18.76 16 1.99
49 Stanislaus 29 34.00 1 5.00 0 0.00 59 2.00 35 4.00
50 Sutter 12 28.10 0 0.00 0 0.00 33 15.80 4 1.80
51 Tehama 19 31.00 0 0.00 3 1.00 15 14.00 9 3.00
52 Trinity 13 52.00 0 0.00 0 0.00 2 0.09 4 0.60
53 Tulare 44 43.00 1 1.00 0 0.00 110 23.00 5 6.62
54 Tuolumne 13 10.28 0 0.00 0 0.00 30 2.86 6 0.00
55 Ventura 27 30.00 B 3 N/A 0 0.00 31 20.00 12 10.00
B
56 Yolo 8 31.47 3 2.59 4 15.72 49 7.26 17 1.87
57 Yuba 6 16.50 1 7.80 0 0.00 10 4.60 17 5.50
Total 1,034 85 125 1,474 923
Average 29.54% 1.23% 0.90 5.42% 4.52%
Legend
A The percentage under school districts includes community colleges, cities, and special districts.
B The percentage under school districts also includes community colleges.
C The percentage of participants for special districts apply to both voluntary and mandatory districts.
N/A Not applicable.
N/P Not provided.
Comments
164
Comments of the California State Auditor on the Response
from the County of Colusa
T
o provide clarity and perspective, we are commenting on the County of
Colusa’s response to our audit report. The numbers correspond to the
numbers we have placed in the response.
. The treasurer’s contention is incorrect. He was contacted by telephone,
and in writing by the audit principal, and was visited by our in-charge
auditor to discuss our comments.
The report to the board of supervisors we received was dated January
30, 1995, not December 16, 1994. In this report, the treasurer states
how he is presently repositioning his portfolio. However, our
recommendations focus on both current and future investment activities since
our report presents data related to the portfolio for two recent points in time
and has relevance to future investment decisions by the treasurer.
The duration calculation measures the interest rate sensitivity for a point
in time—March 31, 1995. Thus, the effect of increases in interest rates
during previous periods is irrelevant.
Unlike the seven other county treasurers we visited, the Colusa treasurer
was unable to provide us with sheets stating the terms of the notes he
had purchased. As a result, we used the confirmation sheets for his
transactions. The weighted average maturity is not an appropriate measure for
the average life of mortgage-backed securities. It is simply the average years
to maturity weighted by the par value of the security. Duration is the appropriate measure to estimate the
average life of the portfolio and interest rate risk. In order to estimate the duration (number of years to
receive the present value of the bond’s cash flow), our experts applied assumptions regarding the
prepayment features of the mortgage-backed securities. To adjust duration for prepayments on
mortgage-backed securities, our investment experts used the PSA Standard Prepayment Model. The
PSA Model was developed by the Public Securities Association, an organization that represents dealers in
mortgage-backed securities. The PSA Model is widely recognized in the investment community as an
appropriate method for adjusting duration for securities with prepayment features. In addition, the
county’s computed effective duration of 24.2 at March 31, 1995, supports the contention that the average
maturity is very long. Finally, our investment experts had a brokerage firm independently compute
Colusa’s effective duration for several of the county’s investments and the results paralleled our
investment experts' computation.
Our investment experts stated that Colusa County’s portfolio was not hedged as of the two dates
analyzed.
Regardless of the issuer, 85 percent of Colusa County’s portfolio consists of securities with
maturities greater than one year that are not actively traded on a secondary market. This was
the largest percentage of the eight counties we reviewed.
Whatever the rationalization, the method does not comply with the allocation method
prescribed by state law.
We do not recommend or suggest a maturity equal to a passbook savings account.
165
Comments
Comments of theCalifornia State Auditor on
the Response from the County of Monterey
T
o provide clarity and perspective, we are commenting on the County of
Monterey’s response to our audit report. The numbers correspond to the
. numbers we have placed in the response.
If the county’s funds are not intended for short-term withdrawal to meet
operational needs, they should not be in the short-term investment pool.
As is done in some other counties, they should be separately invested or
invested in a pool for longer term investors.
The county is incorrect that we used mutual fund data to conclude that
any county’s portfolio is at risk. The use of the mutual fund data was to
acquire relevant information from portfolios with similar objectives and
constraints in order to have a reasonable, but not overly conservative,
comparison group. These included investing in short-term high-quality
investments and the ability to invest in a variety of financial instruments. We
recognize that these mutual funds are not exactly the same as the county pools;
however, these differences are not significant considering how the data was
used, considering the results of our survey of the other counties, and
considering the analysis and recommendations made by consultants hired by
various counties.
While participation in mutual funds is voluntary, the cash flow of these
funds are predictable. For example, the Investment Company Institute (ICI) publishes a report
on trends of mutual fund activity. ICI reports that the flexible bond portfolio (a short-term bond fund
investment category) had historical annual redemptions at approximately 20 percent of the fund’s total
assets. With a 20 redemption rate, 80 percent of the fund’s assets are available for investment in longer
securities. Thus, short-term mutual funds with entirely voluntary participation do not necessarily behave
like that of money market funds which make investment decisions based on having virtually 100 percent
of the portfolio liquid at all times.
166
We do not conclude that Monterey County was imprudent in its investment practices. Our report clearly
indicates that the county’s portfolio is not highly leveraged, nor is its average maturity extremely long.
However, as shown on page 25, it has invested in structured notes with a duration of 2.3, which increased
the interest rate sensitivity of its portfolio above that of its fixed rate note duration of 1.2. However, the
effective duration was the third lowest of the eight counties we reviewed and the effective duration of 2.1
was only slightly higher than our comparison group’s duration of 1.8.
Our report focused on the investment strategies used by Monterey County; thus, we did not
address the techniques that the county did not employ.
The mutual fund data indicated that the average percentage of structured securities was
5.9 percent for those funds reporting structured securities. However, we did not judge
county portfolios using this percentage as the criteria. Rather, we recognized, as stated on page 15, that
all structured securities are not the same and that some structured securities such as certain floaters can
reduce portfolio risk when changes in interest rates occur. Thus, we clearly contrasted San Bernardino’s
use of structured securities which reduced portfolio risk with those of other counties whose structured
notes actually increased portfolio risk. Also, our recommendation does not limit the use of structured
securities to 5.9 percent based on the use of the mutual fund data. Rather, we conclude and recommend
that the use of structured securities that increase market risk to the portfolio should be limited to
5 percent of the portfolio. Increased market risk should be measured by a duration analysis. Although
the county criticizes our use of certain mutual funds, it does not provide any justification for investing in
structured notes that increase risk to the portfolio in excess of 5 percent. Thus, the county has not
established why our conclusion and recommendation is in error.
Our recommendations to the Legislature are intended to assist in deliberating these and other
pending bills.
We did not conclude that the county’s portfolio was because it exceeded 5.9 percent. Our
conclusion is based on the fact that the countyt held 62 percent of its portfolio in structured notes,
which included 33 percent in volatile inverse floaters. The duration of these structured notes was 3.0,
more than twice the duration of the county’s fixed rate notes.
We do not imply that inverse floaters decrease perilously as interest rates rise. However, in
Placer County’s case, our investment experts determined that the county’s inverse floaters, with a
duration of 4.5 at March 31, 1995, contained 3.75 times the interest rate or market risk that its
nonstructured fixed rate notes did on that date, with a duration of 1.2. In addition, our investment
experts analyzed Monterey County’s portfolio as of September 30, 1994, and March 31, 1995, to
determine if the diversification of the county’s investments across alternative rate structures could provide
a hedge against interest rate fluctuations. Their analysis indicated that there is little evidence that
Monterey County hedged its portfolio against interest rate risk on the two dates examined.
The county’s assertion 10 that we have mischaracterized some securities as thinly traded is
incorrect. Our experts attempted to value individual securities held in portfolios and were unable to
readily obtain a market price for many because they are not actively traded on secondary markets and are
thus thinly traded. Furthermore, independent CPAs’ statements in some county audit reports support our
position. Similarly, they were unable to accurately value certain government securities, negotiable
certificates of deposit, and corporate notes because the determination of the value of these investments
involves subjective judgment and can be determined only by negotiation between the parties in a sales
transaction.
Moreover, simply being able to convert a bond to cash does not necessarily indicate that it is free of
liquidity risk. The degree of liquidity risk also is contingent upon the ability to obtain a reasonable price
167
when attempting to sell a security. Because the market for thinly traded securities is limited to the small
number of brokers contacted by the seller, they are subject to greater variation in the bid-ask spread;
namely, at what price can a bond be readily converted to cash. If, however, there is a large spread
between the bid-ask prices, then the bond is said to be less liquid (even though it can be converted to
cash).
168
Blank page inserted for reproduction purposes only.
169
Comments
Comments of the California State Auditor on
the Response from the County of Placer
T
o provide clarity and perspective, we are commenting on the County of
Placer’s response to our audit report. The numbers correspond to the numbers
. we have placed in the response.
Because of legal restrictions related to confidentiality, counties did not
receive data related to other counties.
The county received a copy of the first draft of our report on May 22,
1995, and provided its initial response to us on May 26, 1995.
Therefore, Placer County had our first draft a full two weeks before receiving
our second draft. The county then had over two days to review the changes
that were made and to revise its original response if the county so chose.
As stated on page 44, the audit was performed in accordance with
governmental auditing standards. As the report refers to Placer County,
the facts remain indisputable. The Placer County portfolio contains
64 percent in derivatives known as inverse floaters whose coupon payments
are reduced if interest rates rise. Whether this is called betting,
prognostication, or speculation—no one knows with any certainty the future of
interest rates.
Our audit was confined to the scope outlined on pages 4 and 5, and
conducted in accordance with governmental auditing standards.
We held an exit conference with Placer County on May 24, 1995, at which time we
discussed every issue in the report relating to the county and any conclusions we reached.
The county is incorrect that we used mutual fund data to conclude that any county’s portfolio is
at risk. The use of the mutual fund data was to acquire relevant information from portfolios
with similar objectives and constraints in order to have a reasonable, but not overly conservative,
comparison group. These similarities included investing in short-term high-quality investments and the
ability to invest in a variety of financial instruments. We recognize that these mutual funds are not
exactly the same as the county pools; however, these differences are not significant considering how the
data was used, considering the results of our survey of the other counties, and considering the analyses
and recommendations made by consultants hired by various counties.
While participation in mutual funds is voluntary, the cash flow requirements of these funds are
predictable. For example, the Investment Company Institute (ICI) publishes a report on
trends of mutual fund activity. The ICI reports that the flexible bond portfolio (a short-term bond fund
investment category) had historical annual redemptions of approximately 20 percent of the fund’s assets.
With a 20 percent redemption rate, 80 percent of the fund’s assets are available for investments in longer
term securities. Therefore, short-term mutual funds with entirely voluntary participation do not
170
necessarily behave like money market funds, which make investment decisions based on having virtually
100 percent of the portfolio liquid at all times.
The mutual fund data indicated that the average years to maturity for the securities in the
comparison portfolios was 2.4 years. The county does not believe that the cash flow requirements
of these mutual funds are similar to those of a county, and thus the figure is not relevant to county
portfolios. Our analysis of years to maturity was not performed based entirely on the data from the
mutual fund companies. It also considered, as displayed in Appendix B, that the average years to
maturity reported by 57 counties was 1.4 years. In addition, as displayed on page 21, only 10 of 57
counties exceeded average maturities of 2.4 years during 1994. Furthermore, as stated on page 22,
Standard & Poor’s recommends that the average maturity of county and state investment pools not
exceed one year. Thus, our conclusion, based in part on the data from the mutual funds that the average
maturity of county investment pools should not exceed 2.4 years, is reasonable, not overly conservative,
and would, in fact, only impact 10 of 57 counties during 1994. Moreover, the county has not established
why there is a need for a county’s short-term investment pool to have an average maturity exceeding 2.4
years.
The mutual fund data indicated that the average percentage of structured securities was
5.9 percent for those funds reporting structured securities. However, we did not judge
county portfolios using this percentage as the criterion. Rather, we recognized, as stated on page 15, that
all structured securities are not the same and that some structured securities, such as certain floaters, can
reduce portfolio risk when changes in interest rates occur. Thus, we clearly contrasted San Bernardino’s
use of structured securities, which reduced portfolio risk with those of other counties whose structured
notes actually increased portfolio risk. Also, our recommendation does not limit the use of structured
securities to 5.9 percent based on the use of the mutual fund data. Rather, we conclude and recommend
that the use of structured securities that increase market risk to the portfolio should be limited to
5 percent of the portfolio. Increased market risk should be measured by a duration analysis. Although
the county criticizes our use of certain mutual funds, it does not provide any justification for investing in
structured notes that increase risk to the portfolio in excess of 5 percent. Thus, the county has not
established why our conclusion and recommendation are in error.
As stated on page 14 of 10 our report, 72 percent of Placer County’s portfolio was comprised of
structured notes as of March 31, 1995. Furthermore, as shown on page 124, a Certified Public
Accounting firm hired by the county stated that as of December 31, 1994, approximately 60 percent of the
county’s portfolio was comprised of derivative securities.
We do not imply that all 11 inverse floaters decrease perilously as interest rates rise. However, in
Placer County’s case, our investment experts determined that the county’s inverse floaters, with a
duration of 4.5 at March 31, 1995, contained 3.75 times the interest rate or market risk that its
nonstructured fixed rate notes did on that date, with a duration of 1.2. In addition, our investment
experts analyzed Placer County’s portfolio as of September 30, 1994, and March 31, 1995, to determine if
the diversification of the county’s investments across alternative rate structures could provide a hedge
against interest rate fluctuations. Their analysis indicated that there is little evidence that Placer County
hedged its portfolio against interest rate risk on the two dates examined.
The county’s assertion 12 that we have mischaracterized some securities as thinly traded is
incorrect. Our experts attempted to value individual securities held in portfolios and were unable to
readily obtain a market price for many because they are not actively traded on secondary markets and are
thus thinly traded. Furthermore, independent CPAs’ statements in some county audit reports support our
position. Similarly, they were unable to accurately value certain government securities, negotiable
certificates of deposit, and
171
corporate notes because the determination of the value of these investments involves subjective judgment
and can be determined only by negotiation between the parties in a sales transaction.
Moreover, simply being able to convert a bond to cash does not necessarily indicate that it is free of
liquidity risk. The degree of liquidity risk also is contingent upon the ability to obtain a reasonable price
when attempting to sell a security. Because the market for thinly traded securities is limited to the small
number of brokers contacted by the seller, the securities are subject to greater variation in the bid-ask
spread; namely, at what price can a bond be readily converted to cash. If the sell price (ask) is close to
the buy price (bid), the bond is said to be liquid. If, however, there is a large spread between the bid-ask
prices, then the bond is said to be less liquid (even though it can be converted to cash).
We believe our 13 description of the potential risk associated with collateral calls is
accurate.
If the county’s funds are 14 not intended for short-term withdrawal to meet operational needs, they
should not be in the short-term investment pool. As is done in some other counties, they should be
separately invested or invested in a pool for longer term investors.
Orange County’s 15 investments were liquidated at substantial discounts (up to 25 percent),
thus creating a very active market that would not necessarily have materialized had the securities been
offered at their recorded book values.
Comments
Comments of the California State Auditor on
the Response from the County of Sacramento
T
o provide clarity and perspective, we are commenting on the County of
Sacramento’s response to our audit report. The numbers correspond to the
numbers we have placed in the response.
.
The county is incorrect that we used mutual fund data to conclude that
any county’s portfolio is at risk. The use of the mutual fund data was to
acquire relevant information from portfolios with similar objectives and
constraints in order to have a reasonable, but not overly conservative,
comparison group. These similarities included investing in short-term
high-quality investments and the ability to invest in a variety of financial
instruments. We recognize that these mutual funds are not exactly the same as
the county pools; however, these differences are not significant considering
how the data was used, considering the results of our survey of the other
counties, and considering the analyses and recommendations made by
consultants hired by various counties.
The Government Accounting Standards Board has issued Statement 28,
which will change security lending disclosures for audits of periods
172
ended after December 1995.
Whatever the rationalization, the method does not comply with the allocation method
prescribed by State law.
173
Blank page inserted for reproduction purposes only.
174
Comments
Comments of the California State Auditor on
the Response from the County of San Bernardino
T
o provide clarity and perspective, we are commenting on the County of San
Bernardino’s response to our audit report. The numbers correspond to the
numbers we have placed in the response.
During the week of May 22, 1995, when each county had its draft copy
of the report, we met with or spoke with representatives of each county.
.
In the case of San Bernardino, the treasurer was out of town, so we spoke with
the chief investment officer about his concerns regarding the report. When the
county received its second draft of the report, we again contacted the county.
At this time, we contacted the chief investment officer who relayed the
county’s continuing concerns.
We modified the text of the report. For clarity, on page 15, we footnote
that the structured notes contain these mutual fund investments.
175
Blank page inserted for reproduction purposes only.
176
Comments
Comments of the California State Auditor on the Response
from the County of San Diego
T
o provide clarity and perspective, we are commenting on the County of San
Diego’s response to our audit report. The numbers correspond to the numbers
. we have placed in the response.
The county was provided a copy of all sections of the report that were
applicable to it. Because of legal restrictions related to confidentiality,
individual counties did not receive data related to other counties.
The county received a copy of the first draft of the report on May 22,
1995, and provided its initial response to us on May 26, 1995. Thus, it
had the first draft for two weeks before it received the second draft. It then
had over two days to review the changes that were made and to revise its
original response if it so chose.
The county is incorrect that we used mutual fund data to conclude that
any county’s portfolio is at risk. The use of the mutual fund data was to
acquire relevant information from portfolios with similar objectives and
constraints in order to have a reasonable, but not overly conservative,
comparison group. These similarities included investing in short-term
high-quality investments and the ability to invest in a variety of financial
instruments. We recognize that these mutual funds are not exactly the same as
the county pools; however, these differences are not significant considering
how the data was used, considering the results of our survey of the other counties, and considering the
analyses and recommendations made by consultants hired by various counties.
The mutual fund data indicated that the average percentage of structured securities was 5.9 percent for
those funds reporting structured securities. However, we did not judge county portfolios using this
percentage as the criterion. Rather, we recognized, as stated on page 15, that all structured securities are
not the same and that some structured securities, such as certain floaters, can reduce portfolio risk when
changes in interest rates occur. Thus, we clearly contrasted San Bernardino’s use of structured securities,
which reduced portfolio risk, with those of other counties whose structured notes actually increased
portfolio risk. Also, our recommendation does not limit the use of structured securities to 5.9 percent
based on the use of the mutual fund data. Rather, we conclude and recommend that the use of structured
securities that increase market risk to the portfolio should be limited to 5 percent of the portfolio.
Increased market risk should be measured by a duration analysis. Although the county criticizes our use
of certain mutual funds, it does not provide any justification for investing in structured notes that increase
risk to the portfolio in excess of 5 percent. Thus, the county has not established why our conclusion and
recommendation is in error
177
The mutual fund data indicated that the average years to maturity for the securities in the comparison
portfolios was 2.4 years. The county does not believe that the cash flow requirements of these mutual
funds are similar to those of a county and thus the figure is not relevant to county portfolios. Our
analysis of years to maturity was not performed based entirely on the data from the mutual fund
companies. It also considered, as displayed in Appendix B, that the average years to maturity reported
by 57 counties was 1.4 years. In addition, as displayed on page 21, only 10 of 57 counties exceeded
average maturities of 2.4 years during 1994. Furthermore, as stated on page 22, Standard & Poor’s
recommends that the average maturity of county and state investment pools not exceed one year. Thus,
our conclusion, based in part on the data from the mutual funds that the average maturity of county
investment pools should not exceed 2.4 years, is reasonable, not overly conservative, and would have, in
fact, only impacted 10 of 57 counties during 1994. Moreover, the county has not established why there
is a need for a county’s short-term investment pool to have an average maturity exceeding 2.4 years.
The mutual fund data indicated that the average duration, a calculation measuring the sensitivity to
interest rates, for the comparison portfolios was 1.8. Although the county disagrees with the use of our
comparison group, it is interesting to note that a consultant hired by San Bernardino County performed a
sensitivity analysis for the county and used the Merrill Lynch 1-4.99 year Government Index as a
comparison even though it reflects only government bonds and not other investments purchased by
county treasurers as our comparison group includes. Also, a consultant hired by San Diego County
recommended a duration of only 1.0. Although we present an analysis of duration in our report, we do
not recommend that the duration of county portfolios be limited to any specific amount.
While participation in mutual funds is voluntary, the cash flow requirements of these funds are
predictable. For example, the Investment Company Institute (ICI) publishes a report on trends of mutual
fund activity. ICI reports that the flexible bond portfolio (a short-term bond fund investment category)
had historical annual redemptions at approximately 20 percent of the fund’s total assets. With a 20
redemption rate, 80 percent of the fund’s assets are available for investment in longer term securities.
Thus, short-term mutual funds with entirely voluntary participation do not necessarily behave like that of
money market funds which make investment decisions based on having virtually 100 percent of the
portfolio liquid at all times.
The percentage of structured notes stated in our report is correct. The county’s figures agree to a
summary sheet provided to us. However, the county’s summary sheet excluded certain securities
contained in its portfolio. Specifically, the summary sheet excludes nine structured notes totaling
$328 million.
The county’s assertion that we have mischaracterized some securities as thinly traded is
incorrect. Our experts attempted to value individual securities held in portfolios and were
unable to readily obtain a market price for many because they are not actively traded on secondary
markets and are thus thinly traded. Furthermore, independent CPAs’ statements in some county audit
reports support our position. Similarly, they were unable to accurately value certain government
securities, negotiable certificates of deposit, and corporate notes because the determination of the value of
these investments involves subjective judgment and can be determined only by negotiation between the
parties in a sales transaction.
Moreover, simply being able to convert a bond to cash does not necessarily indicate that it is free of
liquidity risk. The degree of liquidity risk also is contingent upon the ability to obtain a reasonable price
when attempting to sell a security. Because the market for thinly traded securities is limited to the small
number of brokers contacted by the seller, the securities are subject to greater variation in the bid-ask
spread; namely, at what price can a bond be readily converted to cash. If the sell price (ask) is close to
the buy price
178
(bid), the bond is said to be liquid. If, however, there is a large spread between the bid-ask prices, then
the bond is said to be less liquid (even though it can be converted to cash).
Report modified. See page 37.
If the county’s funds are not intended for short-term withdrawal to meet operational needs, they
should not be in the short-term investment pool. As is done in some other counties, they
should be separately invested or invested in a pool for longer term investors.
The issue raised in our report is that the county’s contract does not specifically indemnify the
county if the agent is unable to deliver the security. While the county appears to believe that
industry practice is sufficient protection, we believe explicit indemnification in the contract is warranted
and should not be objectionable to agents if it merely explicitly states industry practice.
Comments
Comments of the California State Auditor on
the Response from the County of Solano
T
o provide clarity and perspective, we are commenting on the County of
Solano’s response to our audit report. The numbers correspond to the
. numbers we have placed in the response.
Our report never suggests that the treasurer intended to put taxpayer
dollars at risk. Rather, similar to the report by the county’s own
consultant, our report recommends that the treasurer reduce risk in the county
portfolio.
The county is incorrect that we used mutual fund data to conclude that
any county’s portfolio is at risk. The use of the mutual fund data was to
acquire relevant information from portfolios with similar objectives and
constraints in order to have a reasonable, but not overly conservative,
comparison group. These similarities included investing in short-term
high-quality investments and the ability to invest in a variety of financial
instruments. We recognize that these mutual funds are not exactly the same as
the county pools; however, these differences are not significant considering
how the data was used, considering the results of our survey of the other
counties, and considering the analysis and recommendations made by
consultants hired by various counties.
While participation in mutual funds is voluntary, the cash flow
requirements of these funds are predictable. For example, the Investment Company Institute
(ICI) publishes a report on trends of mutual fund activity. ICI reports that the flexible bond portfolio (a
short-term bond fund investment category) had historical annual redemptions at approximately 20 percent
of the fund’s total assets. With a 20 percent redemption rate, 80 percent of the fund’s assets are available
for investment in longer term securities. Thus, short-term mutual funds with entirely voluntary
179
participation do not necessarily behave like that of money market funds which make investment decisions
based on having virtually 100 percent of the portfolio liquid at all times.
180
The mutual fund data indicated that the average percentage of structured securities was 5.9 percent for
those funds reporting structured securities. However, we did not judge county portfolios using this
percentage as the criterion. Rather, we recognized, as stated on page 15, that all structured securities are
not the same and that some structured securities, such as certain floaters, can reduce portfolio risk when
changes in interest rates occur. Thus, we clearly contrasted San Bernardino’s use of structured securities
which reduced portfolio risk with those of other counties whose structured notes actually increased
portfolio risk. Also, our recommendation does not limit the use of structured securities to 5.9 percent
based on the use of the mutual fund data. Rather, we conclude and recommend that the use of structured
securities that increase market risk to the portfolio should be limited to 5 percent of the portfolio.
Increased market risk should be measured by a duration analysis. Although the county criticizes our use
of certain mutual funds, it does not provide any justification for investing in structured notes that increase
risk to the portfolio in excess of 5 percent. Thus, the county has not established why our conclusion and
recommendation are in error.
The mutual fund data indicated that the average years to maturity for the securities in the comparison
portfolios was 2.4 years. The county does not believe that the cash flow requirements of these mutual
funds are similar to those of a county and thus the figure is not relevant to county portfolios. Our
analysis of years to maturity was not performed based entirely on the data from the mutual fund
companies. It also considered, as displayed in Appendix B, that the average years to maturity reported
by 57 counties was 1.4 years. In addition, as displayed on page 21, only 10 of 57 counties exceeded
average maturities of 2.4 years during 1994. Furthermore, as stated on page 22, Standard & Poor’s
recommends that the average maturity of county and state investment pools not exceed one year. Thus,
our conclusion, based in part on the data from the mutual funds that the average maturity of county
investment pools should not exceed 2.4 years, is reasonable, not overly conservative, and would, in fact,
only impact 10 of 57 counties during 1994. Moreover, the county has not established why there is a need
for a county’s short-term investment pool to have an average maturity exceeding 2.4 years.
The county’s assertion that we have mischaracterized some securities as thinly traded is
incorrect. Our experts attempted to value individual securities held in portfolios and were
unable to readily obtain a market price for many because they are not actively traded on secondary
markets and are thus thinly traded. Furthermore, independent CPAs’ statements in some county audit
reports support our position. Similarly, they were unable to accurately value certain government
securities, negotiable certificates of deposit, and corporate notes because the determination of the value of
these investments involves subjective judgment and can be determined only by negotiation between the
parties in a sales transaction.
Moreover, simply being able to convert a bond to cash does not necessarily indicate that it is free of
liquidity risk. The degree of liquidity risk also is contingent upon the ability to obtain a reasonable price
when attempting to sell a security. Because the market for thinly traded securities is limited to the small
number of brokers contacted by the seller, the securities are subject to greater variation in the bid-ask
spread; namely, at what price can a bond be readily converted to cash. If the sell price (ask) is close to
the buy price (bid), the bond is said to be liquid. If, however, there is a large spread between the bid-ask
prices, then the bond is said to be less liquid (even though it can be converted to cash).
We are pleased that the Solano County treasurer is following the recommendations even though
he disagrees with the basis for the recommendations.
181
Blank page inserted for reproduction purposes only.
182
Comments
Comments of the California State Auditor on
the Response from the County of Sonoma
T
o provide clarity and perspective, we are commenting on the County of
Sonoma’s response to our audit report. The numbers correspond to the
. numbers we have placed in the response.
We do not conclude that Sonoma County was imprudent in its
investment practices. Our report clearly indicates that the county’s
portfolio is not highly leveraged, nor is its average maturity extremely long.
However, as shown on page 24, it has invested in structured notes with a
duration of 2.3, which increased the interest rate sensitivity of its portfolio
above that of its fixed rate note duration of 1.2. However, the effective
duration was the third lowest of the eight counties we reviewed and the
effective duration of 2.1 was only slightly higher than our comparison group’s
duration of 1.8.
On page 15 of our report, we differentiate between the low-risk floating
notes and other structured notes. Also, our recommendation is to limit
investments in structured notes that increase the risk in the portfolio.
Therefore, it is not intended to limit certain floating rate notes that have very
low durations, and thus do not increase the portfolio’s risk.
While participation in mutual funds is voluntary, the cash flow
requirements of these funds are predictable. For example, the
Investment Company Institute (ICI) publishes a report on trends of mutual fund activity. ICI reports that
the flexible bond portfolio (a short-term bond fund investment category) had historical annual
redemptions at approximately 20 percent of the fund’s total assets. With a 20 percent redemption rate,
80 percent of the fund’s assets are available for investment in longer term securities. Thus, short-term
mutual funds with entirely voluntary participation do not necessarily behave like that of money market
funds, which make investment decisions based on having virtually 100 percent of the portfolio liquid at
all times.
183
The county is incorrect that we used mutual fund data to conclude that any county’s portfolio is
at risk. The use of the mutual fund data was to acquire relevant information from portfolios
with similar objectives and constraints in order to have a reasonable, but not overly conservative,
comparison group. These similarities included investing in short-term high-quality investments and the
ability to invest in a variety of financial instruments. We recognize that these mutual funds are not
exactly the same as the county pools; however, these differences are not significant considering how the
data was used, considering the results of our survey of the other counties, and considering the analysis
and recommendations made by consultants hired by various counties.
The mutual fund data indicated that the average percentage of structured securities was 5.9 percent for
those funds reporting structured securities. However, we did not judge county portfolios using this
percentage as the criterion. Rather, we recognized, as stated on page 15, that all structured securities are
not the same and that some structured securities, such as certain floaters, can reduce portfolio risk when
changes in interest rates occur. Thus, we clearly contrasted San Bernardino’s use of structured securities,
which reduced portfolio risk, with those of other counties whose structured notes actually increased
portfolio risk. Also, our recommendation does not limit the use of structured securities to 5.9 percent
based on the use of the mutual fund data. Rather, we conclude and recommend that the use of structured
securities that increase market risk to the portfolio should be limited to 5 percent of the portfolio.
Increased market risk should be measured by a duration analysis. Although the county criticizes our use
of certain mutual funds, it does not provide any justification for investing in structured notes that increase
risk to the portfolio in excess of 5 percent. Thus, the county has not established why our conclusion and
recommendation are in error.
The mutual fund data indicated that the average years to maturity for the securities in the comparison
portfolios was 2.4 years. The county does not believe that the cash flow requirements of these mutual
funds are similar to those of a county, and thus the figure is not relevant to county portfolios. Our
analysis of years to maturity was not performed based entirely on the data from the mutual fund
companies. It also considered, as displayed in Appendix B, that the average years to maturity reported
by 57 counties was 1.4 years. In addition, as displayed on page 21, only 10 of 57 counties exceeded
average maturities of 2.4 years during 1994. Furthermore, as stated on page 22, Standard & Poor’s
recommends that the average maturity of county and state investment pools not exceed one year. Thus,
our conclusion, based in part on the data from the mutual funds that the average maturity of county
investment pools should not exceed 2.4 years, is reasonable, not overly conservative, and would, in fact,
only impact 10 of 57 counties during 1994. Moreover, the county has not established why there is a need
for a county’s short-term investment pool to have an average maturity exceeding 2.4 years.
The mutual fund data indicated that the average duration, a calculation measuring the sensitivity to
interest rates, for the comparison portfolios was 1.8. Although the county disagrees with the use of our
comparison group, it is interesting to note that a consultant hired by San Bernardino County performed a
sensitivity analysis for the county and used the Merrill Lynch 1-4.99 year Government Index as a
comparison even though it reflects only government bonds and not other investments purchased by
county treasurers as our comparison group includes. Also, a consultant hired by San Diego County
recommended a duration of only 1.0. Although we present an analysis of duration in our report, we do
not recommend that the duration of county portfolios be limited to any specific amount.
The county’s assertion that we have mischaracterized some securities as thinly traded is
incorrect. Our experts attempted to value individual securities held in portfolios and were
unable to readily obtain a market price for many because they are not actively traded on secondary
markets and are thus thinly traded. Furthermore, independent CPAs’ statements in some county audit
reports support our position. Similarly, they were unable to accurately value certain government
securities, negotiable certificates of deposit, and corporate notes because the determination of the value of
184
these investments involves subjective judgment and can be determined only by negotiation between the
parties in a sales transaction.
Moreover, simply being able to convert a bond to cash does not necessarily indicate that it is free of
liquidity risk. The degree of liquidity risk also is contingent upon the ability to obtain a reasonable price
when attempting to sell a security. Because the market for thinly traded securities is limited to the small
number of brokers contacted by the seller, the securities are subject to greater variation in the bid-ask
spread; namely, at what price can a bond be readily converted to cash. If the sell price (ask) is close to
the buy price (bid), the bond is said to be liquid. If, however, there is a large spread between the bid-ask
prices, then the bond is said to be less liquid (even though it can be converted to cash).
185