When to Hold Cash or Defensive Assets

Going to cash is the most expensive insurance policy in finance. From 1928 through 2025, $1 invested in the S&P 500 with reinvested dividends compounded to roughly $8,000. That same dollar in 3-month T-bills compounded to about $23.¹ The opportunity cost of "playing it safe" runs in the high single digits annually — every year. Defensive positioning has a real role in portfolio construction, but the case for it has to clear a high bar, and the implementation has to be disciplined enough that you don't sit in T-bills through the recovery.

This article is about when that bar is actually cleared, what defensive really means, and how to avoid the most common failure mode: defensive on the way down, defensive on the way up.

The Real Cost of Cash (1928–2025)

Damodaran's long-run dataset¹ is the standard reference. Annualized geometric returns from 1928 through 2025:

Cash earns roughly 0.3% per year above inflation over the long run. That is the entire compensation for giving up equity-like upside. Holding 20% of a $500,000 portfolio in cash for a decade costs roughly $130,000 in foregone equity returns, in expectation. That is the number a defensive shift has to "buy" something worth, in either avoided loss or improved sleep.

The point: defensive positioning is not free. Every basis point in cash is a basis point not compounding at the equity premium. Treat it as a paid-for option, not a default.

When the Bar Is Actually Cleared

Three conditions, in order of how robust the evidence is:

1. Valuation Is Stretched and Your Horizon Is Short

The Shiller CAPE — the S&P 500 price divided by 10-year average inflation-adjusted earnings — is the single most-tested long-horizon valuation signal.² Higher starting CAPE has consistently predicted lower 10-year forward equity returns.

(Decile bins from Shiller's published series.³)

For context, CAPE was 13.3 at the March 2009 bottom, about 24 at the March 23, 2020 COVID low, and is currently ~36–40 (April 2026).⁴ Today's reading sits in the top historical decile.

The test: a 65-year-old drawing down a portfolio over the next 15 years and a 35-year-old in accumulation face the same CAPE but very different horizons. Valuation arguments matter most when your investment horizon is shorter than the mean-reversion horizon (10+ years). For a 35-year-old, even an extreme CAPE has historically produced positive real returns over a working lifetime.

2. Cash Yields More Than Inflation, Meaningfully

When 3-month T-bill yields exceed near-term expected inflation by 1.5%+ — as they did in 2023 (T-bills ~5.4% vs. headline CPI falling to ~3%) and through much of 2024–25 — the opportunity cost of waiting collapses.⁵ You're being paid to hold the option to deploy.

Compare to 2010–2015, when Fed funds sat near zero and 3-month T-bills yielded under 0.10%. Holding cash then was a guaranteed loss in real terms. The defensive case had to come from elsewhere (which it generally didn't).

The point: when real cash yields are positive and meaningful, the cost of being patient is small. When real cash yields are negative, the cost of being patient compounds against you.

3. You Have Specific, Near-Dated Liabilities

This is the cleanest case and the most underused one. If you owe a tax bill in March, a tuition payment in August, or a down payment in 18 months, those dollars belong in T-bills regardless of what equities are doing. Liability-driven cash is not market timing. It is matching duration to obligation.

The mistake here is the opposite — investors with imminent liabilities sitting in equity index funds, then forced to sell at a 25% drawdown to meet the obligation. That is a worse error than holding too much cash for too long.

What Defensive Actually Looks Like (Not One Asset, Several)

"Defensive" is not a single decision. Different assets hedge different risks, and they do not substitute for each other.

Cash and T-bills: the universal hedge, the universal drag

Treasury bills are the only asset with no credit risk, no duration risk, and full liquidity. Buy directly from TreasuryDirect.gov⁶ (no fees, no spread) or through a brokerage with auto-roll. State-tax-exempt interest is meaningful for residents of CA, NY, NJ, OR, and similar high-tax states.

A note on money market funds: government MMFs (~95% of MMF assets post-2014 reform) are essentially T-bill proxies. Prime MMFs add corporate paper for ~10–25 bp more yield and very real tail risk — the Reserve Primary Fund broke the buck on September 16, 2008, posting a $0.97 NAV after Lehman's bankruptcy made its $785M of Lehman commercial paper (1.2% of fund assets) worthless.⁷ Within days the Treasury announced a temporary MMF guarantee program and the SEC suspended redemptions. The point: if you're going to use a MMF as cash, use the government version. The yield pickup on prime is not worth the gap risk.

Long-duration Treasuries (TLT): hedge against deflation, not against inflation

iShares TLT calendar-year total returns make the asymmetry obvious:⁸

Long Treasuries are a phenomenal hedge against growth shocks and deflation scares — and a disaster when inflation surprises to the upside. The 2022 result was not a fluke; it is the asset doing exactly what its duration says it should when 10-year yields rise 230 bp in a year.

The 2011 case is instructive on a counterintuitive dynamic: S&P downgraded the US sovereign rating to AA+ on Aug 5, 2011, and the 10-year yield fell from 2.94% on July 28 to 1.72% by late September — a 122 bp rally on the asset that was supposedly downgraded.⁹ Flight-to-quality dominated, even when the quality being fled to was the asset being downgraded.

The durable lesson: long Treasuries hedge equity bear markets driven by recession or financial stress. They do not hedge equity bear markets driven by inflation. Know which scenario you're insuring against.

Defensive sector equities: smaller drawdowns, real upside drag

Utilities (XLU), staples (XLP), and healthcare (XLV) typically run with betas to the S&P 500 in the 0.5–0.8 range, depending on the measurement window.¹⁰ In drawdowns they typically lose less:

(SPDR sector ETF total returns, Bloomberg/Morningstar.)

Two things worth noting: (1) utilities did worse than the S&P during the COVID drawdown — defensive labels do not survive every regime; (2) over the 2010s bull market, XLU and XLP underperformed the S&P 500 by 3–5 percentage points annually. Defensive sectors are not free; you pay in opportunity cost.

Gold: works as a tail hedge, not as a portfolio asset

Gold is volatile, has no cash flow, and decade-long stretches of zero or negative real return are routine (1980–2000 was one such stretch). What it does provide is convexity in monetary or geopolitical regime shifts: 2000–2010 (gold +275%, stocks ~0%), 2008 (+5.5% calendar year while equities lost a third), and 2020 (+25% calendar year). A 5% allocation in a diversified portfolio modestly reduces drawdown variance without meaningfully changing long-run returns; a 20% allocation is a directional bet on monetary instability.

The Full-Bear vs. In-Recession Distinction (Don't Get These Confused)

A common error in defensive-positioning analysis is conflating bear-market drawdowns with recession-window drawdowns. They are not the same thing, and the former is roughly twice the latter.

Equity markets typically peak several months before the NBER recession start and trough before the recession ends. By the time a recession is officially confirmed, most of the equity drawdown has already happened, and the rebound has often started. The implication: waiting for "recession confirmation" to go defensive is waiting until the trade is over.

How Practitioners Actually Use It (Two Real Examples)

David Swensen (Yale Endowment, 1985–2021) ran a relentlessly diversified portfolio, but during the 1999–2000 dot-com peak he allowed Yale's domestic equity allocation to drift down meaningfully via rebalancing — he didn't time the top, but he refused to let valuations carry the weight up. His approach in Pioneering Portfolio Management is explicit: discipline-driven rebalancing toward target weights is the operational form of defensive behavior, not headline-driven tactical shifts.¹³

Ray Dalio's All Weather allocates roughly 30% equities, 40% long-term bonds, 15% intermediate bonds, 7.5% gold, 7.5% commodities — a structural "always-on" defensive posture that aims to be neutral across growth and inflation regimes. The 2022 result (−21% per Bridgewater's reported figures) was a reminder that "all weather" still struggles when both stocks and bonds sell off together.¹⁴

The pattern: practitioners who run defensively do it structurally, not tactically. They commit to an allocation and rebalance into weakness, rather than reacting to news.

The Failure Mode That Kills the Strategy

Every academic study of tactical defensive shifts arrives at roughly the same finding: investors who go defensive systematically under-redeploy on the rebound.¹⁵ The S&P 500 gained 23% in the two months from March 23, 2020 to May 23, 2020. An investor who shifted from 70/30 to 30/70 in early March and stayed there until "things felt safer" missed roughly half of that move.

The protective rule that works is mechanical: predefine the redeployment trigger before you reduce risk. Three options that have empirical support:

1. Calendar redeployment. Commit to spreading any defensive cash back into equities over a fixed window (6 or 12 months) regardless of conditions. Removes the discretion that the data says you'll misuse.
2. Valuation redeployment. If you went defensive at CAPE > 30, commit to redeploying half the cash at CAPE 25 and the other half at CAPE 20, regardless of headlines.
3. Drawdown redeployment. Predefine: at S&P 500 −15%, redeploy 1/3; at −25%, redeploy another 1/3; at −35%, redeploy the rest.

The test, written before you reduce risk: "Under what specific, observable condition will I put this cash back into stocks?" If you can't answer in one sentence with a number, you are going to sit in cash too long.

Detection Signals: You're Misusing Defensive Positioning If…

• Your defensive shift happened after a 15%+ drawdown rather than before one. (You are now buying at the bottom of insurance and selling stocks at the bottom.)
• You can name the trigger that put you in cash but not the trigger that will put you back in stocks.
• Your "defensive" allocation has been 30% cash for more than 18 months, and the original thesis (whichever it was) hasn't been re-examined.
• You shifted to long Treasuries because they "always hedge equities" without checking whether the current equity risk is recessionary or inflationary.
• Your defensive allocation requires selling appreciated positions in a taxable account, and you didn't model the tax cost — which often eats half the expected protection.

Implementation Checklist

Essentials

• Pick the defensive vehicle that matches the risk. Cash hedges everything imperfectly. Long Treasuries hedge deflation. Gold hedges monetary regime shifts. Defensive equities hedge mild drawdowns and offer modest upside drag.
• Cap the size of any tactical shift at 20% of equity allocation. Larger shifts cross from "tilt" into "market timing" and the empirical record on those is poor.
• Write the redeployment rule before you reduce risk. Date it. Sign it.

High-impact refinements

• Run defensive shifts in tax-advantaged accounts where possible. A 35% short-term capital-gains hit on rebalancing in a taxable account often exceeds the protection you're buying.
• For meaningful liability-driven cash (down payment, tuition), use a Treasury ladder matched to the obligation date, not a money market fund. Lock the yield.
• If you use long Treasuries as a hedge, size them on DV01-equivalent terms to the equity exposure you're insuring, not on dollar terms. (See Dollar Duration and DV01 Basics on this site.)

Optional

• A modest gold allocation (3–7%) makes sense if you're worried about specific monetary or fiscal-policy tail risks. It does not make sense as a return-seeking holding.
• Pair defensive shifts with tax-loss harvesting if you have positions sitting on losses. Convert the defensive trade into a tax asset.

Next Step (Run the Math Once)

Take your current portfolio. Add up everything that earns less than the 3-month T-bill rate (cash in checking, low-yield savings, idle brokerage cash). Multiply by the gap between current T-bill yield and what you're earning on it. That is your annual cost of being passively defensive — distinct from intentionally defensive.

For a $400,000 portfolio with $50,000 in a 0.5% checking account and T-bills at 4.5%:

Annual cost = $50,000 × (4.5% − 0.5%) = $2,000 / year

Most people discover their unintentional defensive position is larger than their intentional one. Fix that first, then talk about whether to reduce equity risk on top of it.

---

Footnotes

1. Aswath Damodaran, Annual Returns on Stock, T.Bonds and T.Bills: 1928 – Current (NYU Stern, updated January 2026). Dataset: https://pages.stern.nyu.edu/~adamodar/New_Home_Page/datafile/histretSP.html. Annualized figures computed from posted annual return series. ↩ ↩²

2. Robert J. Shiller, Irrational Exuberance, 3rd ed. (Princeton, 2015). The CAPE methodology divides current real S&P 500 price by the trailing 10-year average of inflation-adjusted earnings. ↩

3. Robert Shiller, Online Data, http://www.econ.yale.edu/~shiller/data.htm. The "ie_data.xls" workbook provides the canonical CAPE series back to 1881. ↩

4. Shiller PE Ratio time series, multpl.com: https://www.multpl.com/shiller-pe. March 2009 reading 13.3; March 23, 2020 reading approximately 24; April 2026 reading in the high-30s. ↩

5. Federal Reserve Economic Data (FRED), 3-Month Treasury Bill Secondary Market Rate (TB3MS), https://fred.stlouisfed.org/series/TB3MS, and CPI series CPIAUCSL. ↩

6. TreasuryDirect, https://www.treasurydirect.gov. Bills available at 4-, 8-, 13-, 17-, 26-, and 52-week tenors. ↩

7. SEC, Litigation Release No. 21025 (May 5, 2009) and SEC, Statement on the Reserve Primary Fund. See also Yale Program on Financial Stability case, United States: Reserve Primary Fund Suspension, 2008, https://elischolar.library.yale.edu/journal-of-financial-crises/. ↩

8. iShares 20+ Year Treasury Bond ETF (TLT), Performance, https://www.ishares.com/us/products/239454/ishares-20-year-treasury-bond-etf. Calendar-year total returns rounded to one decimal. ↩

9. Wikipedia, 2011 United States debt-ceiling crisis, https://en.wikipedia.org/wiki/2011_United_States_debt-ceiling_crisis; CME Group, Debt Ceiling: Lessons from the 2011 Budget Debate (2023), https://www.cmegroup.com/insights/economic-research/2023/debt-ceiling-lessons-from-the-2011-budget-debate.html. 10-year yield trajectory per FRED series DGS10. ↩

10. State Street SPDR Sector ETF fact sheets (XLU, XLP, XLV), https://www.sectorspdrs.com. Reported betas vary with the rolling window; the 0.5–0.8 range covers 5- and 10-year trailing measurements as of late 2025. ↩

11. S&P 500 closing prices via Yahoo Finance / S&P Dow Jones Indices. NBER recession dating: https://www.nber.org/research/business-cycle-dating. ↩

12. S&P Dow Jones Indices, S&P 500 historical drawdowns. Peak-to-trough dating per index closing levels. ↩ ↩² ↩³

13. David F. Swensen, Pioneering Portfolio Management: An Unconventional Approach to Institutional Investment, rev. ed. (Free Press, 2009), Chapters 6–7 on rebalancing discipline. ↩

14. Bridgewater Associates, The All Weather Story (2012), https://www.bridgewater.com/research-and-insights/the-all-weather-story; All Weather 2022 performance reported by Institutional Investor and Bloomberg. ↩

15. See, e.g., Brad M. Barber and Terrance Odean, "Trading Is Hazardous to Your Wealth," Journal of Finance 55:2 (2000), 773–806; Russel Kinnel, Mind the Gap (Morningstar, annual), on the persistent gap between investor returns and fund returns driven by mistimed reallocation. ↩