Performance Analytics · The complete picture
"Performance" is not one number. For a public fixed-income pool it is a dozen-plus distinct measurements, yield, total return, duration, mark-to-market, risk-adjusted, peer-relative. Here is every lens the data supports, every one it doesn't, and why the difference matters.
The complete metric catalog
Every performance, risk, liquidity, credit, and income metric used to evaluate a pool like this, with Alameda's actual value where the public data allows, and an honest flag where it doesn't. The gaps are not ours; they are what the County's reporting omits.
| Data? | Metric | What it captures | Alameda |
|---|
Lens 1 · Return, five ways
Book yield, yield-to-maturity at market, the cash-basis effective rate of return, and the stated benchmark all describe the same portfolio, and diverge by more than a full point in places. Which one a reader sees shapes the entire story.
Fiscal year-end. All in %. The County leads with book yield; total return, the economic figure, is not published.
Lens 2 · Book vs. market, what book yield hides
Subtracting book value from market value reveals the unrealized gain or loss the book-yield report never shows. As rates spiked, the pool's market value fell $204M, $235M, and $109M below cost in FY2022–FY2024, an economic reality invisible in a 3–4% book yield. It has since recovered to a small gain.
Bars: market value minus book value ($M). Line: market ÷ book (a "price" of the pool).
To its credit, the County reports unrealized gain/loss and NAV. But it never combines price change with income into a total return. Book yield read a placid 3–4% while the pool's NAV ratio sat at 0.976 in FY2022–23, worth ~2.4% less than its carrying value. Total return reconciles the two; no peer pool publishes it.
A reported total return would have shown the FY2022–23 dip and the recovery. Book yield alone showed neither.
Lens 3 · Benchmark-relative
The County's benchmark variance was positive for three years, then −54 and −84 basis points in FY2023–FY2024. The monthly report keeps plotting book yield against ICE BofA indices, but it compares book yield, not total return, and against Treasury indices that don't match the holdings. A proper evaluation re-runs this on a total-return basis against a duration-matched index.
Book yield minus the stated benchmark, basis points (FY2020–FY2024 shown).
Lens 4 · Risk & risk-adjusted return
Reward should be judged against risk, but read the screen carefully. Alameda carries the longest weighted-average maturity of its cohort (842 days), and a simple yield-per-year-of-duration ratio makes it look under-rewarded. That reading misleads: the pool extended through the near-zero-rate era, so its long sleeve is partly the low-coupon legacy any long book would carry, yet it still earns a solid 4.18%, propped by high-coupon callables bought near the 2023 peak. The sharper questions are whether that yield is durable (those callables can be called away) and whether the position is measured on the right basis (total return), not whether it is "under-rewarded" for duration.
Each point is a fiscal year: duration (WAM, days) vs. book yield. The path shows the pool extending maturity through the rate cycle.
Book yield ÷ WAM-in-years, return earned per unit of interest-rate risk.
A closer look, the yield puzzle
A pool that has run long for years should be weighed down by old, low-coupon bonds bought in the 2020–21 zero-rate era, so its book yield should sit below 4.18%. It doesn't. That single anomaly is one of the most important questions an evaluation can answer: is the yield durable, or about to roll off? It breaks down two ways.
High-coupon callable agencies. Roughly half the pool is agencies, many callable, bought near the 2023 rate peak at 5%+ coupons. They lift book yield and stretch stated maturity, the longest WAM of any large CA county pool (842 days), but they will likely be called.
Timing & selection. The pool did extend and lock in higher coupons. Repeatable skill, or one good cycle? A fair, and testable, question.
The recast shows the tell: to-maturity duration ~2.1yr vs. the County's option-adjusted ~1.55yr. That ~0.55yr gap is the callable optionality inflating the "long" look.
↓ Call & reinvestment risk. As rates ease, the high-coupon callables get called and the cash reinvests lower, dragging book yield down.
↑ Roll-up. The legacy 2020–21 zero-rate bonds mature and reinvest higher, lifting book yield.
Which force wins is the difference between a yield that holds and one that fades, what a book-yield snapshot can't tell the Board, and what this evaluation will.
Resolving it to the dollar requires the lot-level purchase yields and call schedules in our data request, the "last mile" beyond public data.
Lens 5 · Interest-rate sensitivity, liquidity & income
Est. market-value change at ±100 / ±200 bps (duration ≈ 2.0 yr × $10.6B)
From the pool's reported effective duration (1.55 yr at 3/31/26) × value. Agency-callable convexity would dampen the upside; precise key-rate impacts need holdings-level data.
Share maturing in each band (FY2024 audited)
FY2024 total investment income, $M
Lens 6 · The recommendation, in action
Here is the missing number, reconstructed: the pool's estimated total return (income + price change) against the ICE BofA 1–3yr U.S. Treasury index. Total return tells the honest story that book yield hides. The pool went underwater through the 2022–23 rate selloff, then healed as its bonds pulled toward par, and over the full cycle the total-return and book lines converge. That full picture, the drawdown and the recovery both visible, is the reporting the County should adopt.
Annual %, fiscal year. Total return and the ≈fiscal-year index are HSG reconstructions from the County’s public reports. The full-cycle total return and the FY22–24 drawdown are robust; the single-year price splits (notably across the 2022–23 selloff) are preliminary, and we validate them exactly against the County’s security-level files in the engagement. The recommendation is for the County to compute and report this directly.
The bottom line
Of the forty-one ways to measure this pool, the County's reporting is actually relatively strong, it publishes book yield, effective duration, WAM, unrealized gain/loss, NAV, and credit quality. The decisive gap is the one metric no peer pool publishes either: a mark-to-market total return against a named index. Closing it is the heart of the recommendation.
Every measure on this page is reproducible from the public record. The County's report shows a handful; an independent evaluation shows them all, and reconciles them.
← Back to the evaluation