Independent Evaluation · RFP 902732 · FY2020–FY2026
A six-year, source-verified review of the Treasurer's Investment Pool, performance, composition, governance, and reporting, built to answer the question the Board actually convened it for: can the proposed Ethical Investment Policy move forward without hurting returns?
The pool at a glance
Sources: Alameda County Treasurer monthly investment reports (Mar 2026) and FY2024 audited financial statements (MGO CPAs). Policy WAM cap of 3 years (1,095 days) per the 2024 Investment Policy.
Performance & benchmark
The pool's book yield climbed from under 1% to 4.18%. But book yield is an accounting figure: it trailed the County's own benchmark by 54–84 bps in FY2023–FY2024, and it can't show the mark-to-market reality underneath. The number that would show it, a true total return, is published by neither Alameda nor any peer pool we surveyed.
Fiscal year-end (June 30). Yields in %. Benchmark = the County's stated ICE BofA Treasury composite; values plotted here FY2020–FY2024.
To its credit, Alameda reports relatively well, book yield, yield-to-maturity, effective duration, weighted-average maturity, unrealized gain/loss, NAV, and credit quality. But across all seven peer pools we surveyed, not one publishes a true total return, income plus price change, against a named index. What's labeled "rate of return" is a book-basis figure. CDIAC itself draws the line between a book-yield portfolio (what Alameda runs) and a total-return portfolio; the economic number the Board needs to see is the one that's missing.
This is the single largest gap between public-pool reporting and investment-manager-grade reporting, and exactly what an independent evaluation should close.
Growth of $100 invested January 2020, from 75 monthly reports recast through March 2026. Book return = the County's reported income return; total return = income plus the monthly mark-to-market price change, chain-linked.
The two lines converge over the full cycle, vindicating hold-to-maturity, but the total-return line spent nearly three years underwater (2022–2024), bottoming at an estimated −4% NAV in September 2022 (a public-data estimate, to be finalized against County records; every short, high-grade public pool sat below par in that hiking cycle), while book return showed only smooth gains. That gap is the rate risk book-yield reporting hides, and it matters most when participants can withdraw at book. The same month-by-month series drives our downloadable Excel recast.
A pool that has run long for years should carry old, low-coupon zero-rate-era bonds dragging its book yield below 4.18%, yet it sits right at 4.18%. Two things hold it up (high-coupon callable agencies bought near the 2023 rate peak, plus genuine timing), and two forces decide where it goes next (calls & reinvestment pulling it down as rates ease; legacy bonds rolling up). Whether it holds or fades is exactly what a book-yield snapshot can't tell the Board, and what this evaluation answers.
See the breakdown, the yield puzzle →We thought that was the central risk, and built a full run-risk simulator to test it. Then we looked closer: we pulled all 14 Alameda cities' audited financials and found the cities aren't in the pool at all, they self-invest through LAIF. The voluntary half is the County's own funds, and the rest are captive districts that can't leave, so the run-risk is far lower than it first appears. And that very fact turns out to explain why the pool never reports total return, there's no outside investor to demand it. The simulator that explored it, the investigation that overturned it, and the full workbook all sit inside the gated deep analysis.
See what we found, the investigation, the simulator & the workbook →Total return is the metric CDIAC calls "the most ideal," that neither Alameda nor its peers publishes. We didn't propose to build the capability; we built and tested it. It ingests the five standard SymPro/custodial exports, strips the County's own cash flows, and returns an auditable total return. Validated end-to-end on production-format data, it reproduced our public-data recast to the penny (+17.23%). On day one of data access, it runs on your uploads.
See the engine, five inputs, one auditable number →Portfolio composition
Federal agencies peaked at 58.3% of the pool in FY2024 and have eased to 49.5% by March 2026 as corporates rose to 16.9% and new sleeves (CMBS/MBS, a fixed-income JPA) appeared. Weighted-average maturity extended from a defensive 473 days (FY2023) to 842 days, still well inside the 3-year policy cap.
Source: Alameda County Treasurer monthly investment reports, FY2020–Mar 2026.
% of pool by fiscal year-end
% of pool by issuer, audited
Concentration & credit
As of June 30, 2024, Federal Home Loan Bank alone was 45.98% of the pool, policy-compliant (agencies may run to 100%), but a textbook diversification discussion. By March 2026 that single-issuer concentration had eased to ~25% as the book diversified. Credit quality is uniformly high (AA+/Aaa agencies, A-or-better corporates), consistent with a safety-first mandate.
Source: FY2024 audited financial statements, Note 3 (concentration of credit risk).
National benchmarking
We assembled a source-verified comparison set: every major California county pool plus eleven of the largest state and local government investment pools nationwide. Two things stand out: Alameda runs one of the longest durations of any large county pool, and it is one of the few without a consistent, market-based benchmark in its public reporting.
Each bubble is a county pool: horizontal = weighted-average maturity (days); vertical = reported book yield (%); bubble area ∝ assets under management. Latest available reports.
| Government investment pool | State | AUM ($B) | Yield % | Rating | Manager | ESG policy |
|---|
Alameda (highlighted) is a longer-duration county operating pool, not a money-market LGIP, so its higher book yield reflects the extra duration and credit risk it carries, not outperformance. Yields are the latest published 7-/30-day or book yields and are not strictly like-for-like across pool types; shown for scale. Sources: each pool's official fact sheet / disclosure and S&P Global Ratings.
Reading the findings
Two of these findings carry real consequences for the Board, so here's each one explained twice: once for a decision-maker who isn't a markets specialist, and once for the finance professionals in the room.
Book yield is the interest the pool's bonds were promised to pay, measured against what the County paid for them, an accounting figure that barely moves. Total return is what the portfolio actually earned or lost: that interest plus the change in what the bonds are worth today. Reporting only book yield is like judging a rental property by the rent check while ignoring whether the building's value rose or fell, it can look reassuringly steady even while the underlying value is dropping. When interest rates jumped in 2022, the pool's holdings were worth less than the County paid, underwater by more than $200 million for three straight years, yet book yield kept showing a smooth, positive number that revealed none of it (or the recovery since). The Board has been steering with a gauge that can't see the bumps; if the County ever had to sell before maturity, book yield would give no warning of the loss it would take.
Book yield is an amortized-cost income yield, accrual accounting that lags the market and ignores price. Total return (income + price change) is the economic performance measure: what every investment manager reports and what GIPS requires. Alameda discloses the components, book yield, unrealized gain/loss, NAV, but never combines them into a total return against a duration-matched index. Without it you cannot measure risk-adjusted performance, cannot benchmark like-for-like, and cannot tell whether the duration and credit risk in the book is actually being rewarded. CDIAC itself separates a "book-yield portfolio" from a "total-return portfolio": book yield is fine for budgeting income, but it is not a performance metric.
The pool keeps its money invested longer than any other large California county pool, and here's the part that's easy to miss: it built much of that long position during the near-zero-rate years, when long bonds paid almost nothing. A book like that should be weighed down to a low yield today. Yet it still earns a solid ~4.18%, that's to the Treasurer's credit, not a knock. So the useful questions aren't "why so little reward for the risk" (the pool is actually earning well); they're two others. First, is that yield durable? Much of what holds it up are high-coupon bonds bought near the 2023 rate peak that can be called away and reinvested at lower rates. Second, can the Board even see whether the long position is working? The one number that shows it, a true total return, is the number the pool doesn't report. The duration is a coherent bet that rates fall, and lately it has paid off; the risk doesn't vanish because the County owns the money; a fiscal emergency, or a mandated Ethical-Investment-Policy divestment, could still force a sale of an underwater bond (as with the ~$32M Caterpillar divestment in late 2024–early 2025). So the Board should be able to watch the bet on the right gauge, and decide whether to keep making it.
The pool carries the longest effective duration (≈1.55 yr) and WAM (842 days) of the eight-county cohort, and on a simple yield-per-unit-of-duration screen it can look under-rewarded. Read carefully, that screen misleads: the low ratio largely reflects extending through the zero-rate era, low-coupon long paper any such book would carry, not weak management. The 4.18% book yield is, in fact, strong given that legacy, propped by high-coupon callables bought near the 2023 peak. The substantive issues are three: (1) durability, call/reinvestment risk as those callables roll off; (2) measurement, the position can only be judged on a total-return basis versus a duration-matched index, which the pool doesn't report; and (3) composition, incremental safe yield comes from sector/spread within §53601, not from more length. The rate risk is real (≈ ±$164M per 100 bps on the ~$10.6B book) and a forced sale (an operating-liquidity shock or a mandated EIP divestment) would crystallize it, so the duration should be measured and monitored, not reflexively cut.
These findings connect: the pool is making a deliberate interest-rate bet (Finding 2), but the only honest way to know whether that bet is winning or losing is the very number that isn't being reported (Finding 1). Fix the reporting and the strategy becomes a transparent, monitored decision the Board can own; leave it, and a large rates position keeps running without a scoreboard. Neither is a sign of a badly-run pool (Alameda is well-managed and reports above its peers), but both are exactly what an independent evaluation exists to surface.
Governance & advisory structure
For California county pools above ~$10B, in-house management with a statutory Treasury Oversight Committee and an annual independent audit is the dominant model, Los Angeles, San Diego, Riverside, San Bernardino, Sacramento, and Alameda all manage in-house. External advisors appear in only two cases in our set: Santa Clara (Meeder) and Orange County, and Orange only after a 2024 governance crisis transferred investment authority out of the Treasurer's office.
The entire California oversight framework, Gov. Code §§27130–27137, exists because of Orange County's 1994 bankruptcy. That history is the reference point for any governance recommendation here.
Sources: county investment reports & ACFRs; Voice of OC (Feb 2025); CA Gov. Code §§27130–27137. Advisory fees typically 10–35 bps.
The benchmark question
To its credit, the County does show a benchmark: the ICE BofA 0-3yr and 0-5yr US Treasury indices, plotted against the portfolio monthly since 2015. Three things undercut it as a true performance yardstick: it's shown as 18- and 30-month moving averages (which can't be held as a portfolio), the Treasury-only indices don't match an agency- and corporate-heavy pool, and the County compares its book yield to them rather than total return against total return.
"The County does not manage the portfolio to a specific benchmark; rather, benchmarks are used to show the County is earning a market rate of return over a period of time."Source: County of Alameda monthly investment report, March 2026, p.4 (verbatim)
GFOA, the CFA Institute (SAMURAI), and GIPS all call for a benchmark that is investable, composition-matched, total-return-based, and one the manager is accountable to. A market-grade benchmark is not a "reporting-discipline" virtue badge; it is what a fund relies on when it actively manages a duration position and needs to see whether the bet is working. The clearest California example is the City of San Diego: it splits its pool into a liquidity sleeve and a longer "core" (0–5yr) managed and reported on a total-return basis against the ICE BofA 1–3yr U.S. Treasury index. Most California county pools, by contrast, still benchmark book yield against LAIF's apportioned yield, a yield-to-yield comparison, not a total-return one.
We pressure-tested this, it survives the obvious objections
Two objections could dismiss the San Diego comparison, and neither holds. "They report total return because participants can run on them, forcing a market valuation." The opposite is true: San Diego's pool is 100% the City's own money, no external participants who could ever run, the same low-run-risk feature we found in Alameda's voluntary base. It marks to market with no run pressure at all. "They only report total return because it always flatters them." Also false: when rates rose, San Diego realized capital losses and posted negative yields; it reported the bad, not just the good.
So the driver is neither run-risk nor optics; you cannot govern a duration position you don't measure. The honest read for Alameda: low run-risk is a legitimate reason book-yield accounting hasn't created an equity problem between participants, but it is not a reason to fly without the instrument. A comparably low-run-risk, all-internal pool reports total return anyway, purely so its board can see whether the rate bet is paying. And because our reconstruction shows Alameda's total return is sound, adopting it is a governance upgrade the County can claim at no cost.
Sources: County of Alameda monthly report (Mar 2026, p.4); GFOA "Using Benchmarks to Assess Portfolio Risk and Return"; CFA Institute & GIPS 2020; CDIAC reporting guides; City of San Diego Pooled Investment Fund, 100% City funds; Liquidity + Core (0–5yr) structure, Core managed to total return vs. ICE BofA 1–3yr U.S. Treasury; realized capital losses / negative yields reported during the 2022–23 rate rise (sandiego.gov/treasurer, investment FAQs & monthly reports).
One of the central questions behind this review
This is one of the central questions the County convened this evaluation to weigh, alongside an independent read of the Pool's performance, benchmarking, governance, and reporting, each valuable in its own right. In October 2025 the Board approved an Ethical Investment Policy and then froze it, directing that it not take effect until an independent peer review reports back on its impact. That review is this engagement. On the EIP, the decision turns not on whether the policy is morally right, but on whether it can be implemented without hurting returns or performance. The screens target issuers earning more than 10% of revenue from oil/gas/coal, firearms, tobacco, casinos, correctional facilities, distillers, or defense, plus a human-rights prohibition, among the most expansive ethical screens ever proposed for a U.S. public operating pool.
The House Strategies recommendation
Never sell a screened bond at a loss, and the EIP can never lock one in.
A bond is a promise to repay its full face value on a set date. Its price can dip below what we paid when rates rise, but it always climbs back by maturity. That gives one simple, loss-proof rule for divesting any screened holding:
Above the line (worth at least what we paid) → sell now: a clean exit at no loss, often a small gain we reinvest at today’s higher yields. Below the line (temporarily underwater) → don’t sell; let it mature, when it repays par, then move the cash to a compliant bond. Nothing here is more than ~3 years from maturity, so every underwater name floats back within 3 years. We sell the winners now and mature out the dips, divesting the whole screened list without ever locking in a loss. (And prohibited holdings are $0, screened-industry exposure 0.01% today, so there is almost nothing to sell in the first place.)
Share of the pool exposed to the EIP's sector screens
~90% of the pool (U.S. Treasuries, federal agencies, supranationals, and bank CDs) is structurally outside the screened sectors. The policy bites only on the corporate slice (~$1B). Treasurer Levy stated he does not expect a lower yield.
Operating-fund precedent is vanishingly rare
Across all 50 state LGIPs and every county pool we surveyed, formal ESG/ethical screening of an operating pool exists in just three places, and where measured, the return impact was reported as negligible.
| Entity | Adopted | Reported impact |
|---|
Sources: LA County, Multnomah County, City of Portland policies; Oaklandside/KQED (Oct 2025); Friede-Busch-Bassen (2015); MSCI; World Bank/GPIF (2018).
Innovation & forward practice
An independent evaluation shouldn't stop at "is the pool well-run?" It should ask "what does leading practice look like now, and what's within reach?" We surveyed the most innovative practices in public fixed-income today, each already in use at a named public fund, and mapped every one to Alameda's exact mandate: safety first, ≤5-year maturity, investment-grade only, no leverage or derivatives, fully within Government Code §53601.
| Innovation | Public funds already doing it | Fits §53601, no added risk? |
|---|---|---|
| Two-portfolio structure, a liquidity sleeve plus a longer "core" managed for total return | Fairfax County (VA); City of San Diego; Oregon Short-Term Fund | Yes, structure only |
| Total-return management vs. a market index, the core measured on income + price against a duration-matched benchmark | San Diego (vs. ICE BofA 1-3yr Treasury); Palo Alto; Chandler-managed CA agencies; Florida Intermediate Duration | Yes, measurement only |
| GIPS-compliant, total-return performance reporting | CalPERS (asset-owner GIPS-verified, incl. its §53601 STIF); Pennsylvania Treasury (first U.S. state treasury); Chandler (verified since 1997) | Yes, reporting standard |
| Weekly shadow-NAV & liquidity-coverage stress testing (S&P AAAm template) | Virginia LGIP; 40+ S&P-rated pools (TexPool, COLOTRUST, Florida PRIME) | Yes, analytics overlay |
| OAS / key-rate-duration analytics on the callable-agency book | GFOA best practice; platforms at Chandler, Clearwater for Pooled Funds, BlackRock Aladdin | Yes, analytics only |
| Public open-data transparency dashboard | Illinois Treasurer ("The Vault"); Washington State LGIP, and this portal | Yes, disclosure only |
| Sustainability tilt without divestment, Sustainalytics scoring as a tiebreaker; labeled green/social supranationals | LA County; Santa Clara; San Diego; San Francisco; CA PMIA (World Bank green bonds); Nuveen impact IG | Yes, within IG / §53601(q) |
| Community-impact & linked deposits at local banks, MDIs, and CDFIs (collateralized / FDIC-insured) | SF Treasurer local-bank CD program; CA State Treasurer Time-Deposit / CDFI program | Yes, §53601(f), collateralized |
Sources: each fund's official disclosures, fact sheets, and policies; GFOA "Managing Market Risk in Investment Portfolios"; S&P Global AAAm surveillance; CDIAC; CFA Institute / GIPS 2020. Full citations in the engagement report.
Six moves that advance performance, transparency, and the County's stated values without changing its safety-first, ≤5-year, investment-grade risk posture. Each is an option for the Board to weigh, not a criticism of a pool that already reports above the norm.
Carve the pool into a liquidity sleeve sized to participant cash needs and a "core" sleeve managed and measured on total return vs. the ICE BofA 0-3yr Treasury index. Fairfax and San Diego run exactly this. An investment-policy amendment, no new instruments, no higher duration ceiling.
Report a mark-to-market total return against a duration-matched index each quarter, and move toward GIPS-aligned presentation, following Pennsylvania Treasury and CalPERS. Closes the single biggest reporting gap we found, at zero portfolio cost.
Adopt the Virginia LGIP / S&P template, a weekly market-value NAV plus a published "days-of-liquidity under stress" coverage figure. A pure analytics overlay that strengthens the safety-and-liquidity mandate.
Pair the EIP's negative screens with a positive tilt, labeled green/sustainable supranationals inside the existing IBRD/IFC allocation (the CA PMIA model) and collateralized deposits at local / MDI / CDFI banks (the SF / CA-STO model). Advances the County's values at no credit or duration cost.
Productize this very portal, machine-readable holdings, total-return-vs-index, and a real-time EIP screening & tracking-error monitor. Illinois's "Vault" sets the transparency precedent; a live EIP-impact tracker would be genuinely first-in-nation.
Apply option-adjusted-spread and key-rate-duration analysis to the callable-agency holdings (GFOA best practice; standard in Clearwater / Aladdin). Sharper risk capture and better relative-value execution, same instruments, same limits.
None requires a new asset class, more duration, leverage, or any departure from investment-grade safety. They are innovations in structure, measurement, transparency, execution, and impact, the dimensions where a well-run book like Alameda's can still move from good to leading. Surfacing exactly that is the value of an independent evaluation.
Investment strategy, beyond measurement
The Pool is already the longest-duration pool in California; the real question is whether it's the strongest. Today it isn't: it earns only a middling yield for the most duration in the state. The levers below close that gap, more return for the risk already carried, using only instruments already permitted by California Government Code §53601 and the County's Investment Policy (investment-grade, ≤5-year, no leverage, no derivatives), keeping Safety → Liquidity → Yield, in that order.
Manage effective duration to the rate cycle inside a stated band (e.g., 1.25–2.25 yr) instead of letting it drift, extend when the curve is steep and the Fed is easing, shorten when rates rise. The Pool's single biggest total-return lever, and the direct answer to its longest-duration posture.Trade-off: needs a rate view; the band caps how wrong you can be.
Sit at the steepest part of the 0–5 year curve so bonds earn a price gain as they "roll down" and age, and choose barbell / bullet / ladder to fit the curve and the rate view.Trade-off: roll-down shrinks if the curve flattens or inverts.
Rotate among the permitted sectors (Treasuries, agencies, supranationals, IG corporates, high-grade ABS) toward whichever offers the best option-adjusted spread, earning incremental yield at the same credit quality.Trade-off: needs active monitoring; modest spread volatility.
Overweight AAA/AA supranationals (sovereign-quality credit, extra spread) and add short, high-grade ABS / agency CMBS within the §53601 caps, yield without real credit risk.Trade-off: prepayment & complexity, size modestly.
Callable agencies pay more because you've implicitly let the issuer call them; buy them only when the option-adjusted spread compensates, with laddered call dates and a capped allocation.Trade-off: extension / reinvestment risk, the riskiest lever; OAS discipline required.
The floor analysis settles the "longest-duration-in-California" worry, and turns it into the opportunity. First, the length is justified: the County's own flows show only ~33% (~$3.5B) ever needs to be liquid, so a ~$7B base can sit long. Second, and counter to the "too-long" read, the pool is actually running ~$0.4B thin on that liquid buffer, not over-extended. So the problem was never the length; it's that the longest pool in the state earns only a middling yield. The move isn't to shorten; the goal is to be the strongest: earn more for the duration already carried, through composition (levers 1–5: spread, roll-down, total-return management), and top the liquidity buffer up modestly.Top up the ~$0.4B buffer; lift safe yield through composition (spread and roll-down), not more duration.
See the floor analysis on the County's own data →Across dozens of participants, one's outflow date often lands near another's inflow, so the net county-wide cash need is far smoother than the gross. Size the liquidity sleeve to the net (plus a stress buffer), freeing more of the book to safely earn. The gross per-participant flows aren't public (only the net is), so we quantify this in execution, from the County's apportionment ledger (a first-phase task).Trade-off: model flow correlation; keep a stress buffer.
No new asset class, no higher duration ceiling, no leverage, no derivatives, only the permitted §53601 instruments, managed more actively. These are return levers on the investment side (how the money is put to work), not measurement, each an option for the Board to weigh against its own risk tolerance. The first five are standard tools at sophisticated public-fund managers; the last two are original to Alameda's structure.
How House Strategies Group would deliver
This portal is a preview of the work itself, every claim sourced, every metric traceable. Our engagement turns it into a complete, defensible evaluation.
Six+ years of holdings, yields, and reporting, independently rebuilt from primary sources, not the County's summaries.
Total-return performance against a duration-matched, investable index, plus a national peer cohort by size, structure, and objective.
Oversight structure, advisory model, and the transparency of what reaches the Board, measured against GFOA and CDIAC best practice.
Quantify the EIP's real reach and return impact against the actual book, with screening, monitoring, and implementation methodology.
A written report and a Board of Supervisors presentation, practical, implementable, and built to withstand public scrutiny.
Built from the County of Alameda’s public monthly reports as a capability demonstration for the RFP 902732 evaluation. Figures derived here are estimates pending validation against County records. At procurement close this site converts to a private engagement workspace.
Prepared by House Strategies Group in support of County of Alameda RFP 902732, Treasury Pool Evaluation Services. Fiduciary, independent, and grounded entirely in the public record.
Review the analysis ↑