Standard asset pricing theory tells a clean story: an asset's price reflects the discounted stream of its expected cash flows, adjusted for risk. Yet a persistent anomaly haunts this elegant framework. Treasury bills, high-grade sovereign debt, and other safe assets consistently trade at yields far below what their fundamental characteristics would predict.

This gap, often called the convenience yield or liquidity premium, has grown from a curious empirical footnote into one of the most consequential puzzles in modern macroeconomics. Krishnamurthy and Vissing-Jorgensen have documented convenience yields on U.S. Treasuries averaging roughly 73 basis points historically, with spikes during periods of financial stress that reveal something deeper than mere flight-to-quality dynamics.

The implications extend well beyond academic curiosity. If safe assets provide non-pecuniary services—collateral value, settlement finality, informational insensitivity—then their pricing carries information about the state of financial intermediation itself. For monetary policymakers navigating an environment of persistent safe asset scarcity, understanding these premia is not optional. It determines whether interest rate policy transmits as intended, whether quantitative easing operates through portfolio balance channels or scarcity channels, and whether sovereign debt management should be understood as fiscal policy or as public liquidity provision. The puzzle, in short, sits at the intersection of asset pricing, monetary economics, and public finance.

The convenience yield framework, formalized most rigorously by Krishnamurthy and Vissing-Jorgensen, posits that certain assets deliver a stream of non-pecuniary services that command a premium beyond their cash flow characteristics. Treasuries, in particular, function simultaneously as stores of value, collateral in repo markets, settlement instruments in derivatives clearing, and informationally insensitive claims that require minimal due diligence in transactions.

Consider the mechanics: when a Treasury security serves as collateral in a triparty repo transaction, it enables leveraged trading that would otherwise be impossible or prohibitively expensive. The security's owner captures both the yield and the collateral service value. This dual functionality creates a wedge between the observed yield and what a pure discounted cash flow model would predict.

The empirical signatures of convenience yield are striking. The spread between Treasury yields and matched-maturity synthetic equivalents constructed from interest rate swaps—the so-called Treasury-swap spread—turns negative in ways that pure credit risk cannot explain. Nagel's work on the near-money premium demonstrates that shorter-maturity Treasuries command especially large convenience yields precisely because they most closely substitute for money in institutional balance sheets.

Importantly, convenience yield is not static. It expands dramatically during periods of financial stress, reflecting heightened demand for informationally insensitive collateral. During the March 2020 Treasury market dislocation, we observed the paradoxical unwinding of these premia as forced selling overwhelmed the convenience demand—a reminder that even the safest assets have finite absorption capacity.

The theoretical implication is that asset pricing must accommodate a moneyness dimension beyond traditional risk factors. Assets differ not only in their expected returns and covariances but in the transactional and collateral services they render. This reshapes how we think about the risk-free rate itself.

Takeaway

The risk-free rate is not risk-free—it is service-laden. Safe assets are priced as if they were partially money, and the premium reflects the value of that monetary quality.

The global demand for safe assets has grown structurally faster than supply for at least two decades. Caballero, Farhi, and Gourinchas have developed a compelling framework in which emerging market savings, aging demographics in advanced economies, and regulatory demand from banks and pension funds have generated a persistent excess demand for informationally insensitive claims.

The supply side has failed to keep pace. The Eurozone crisis reclassified peripheral sovereign debt from safe to risky in a matter of months, contracting the effective global supply. Basel III liquidity coverage ratios, EMIR collateral requirements for derivatives clearing, and money market fund reforms have all mechanically increased institutional demand for the shrinking pool of AAA-rated sovereign debt.

This imbalance manifests in depressed real yields on government securities across advanced economies. The natural rate of interest—r-star—has fallen not only because of demographic and productivity headwinds but also because the marginal saver has been pushed into safe assets whose supply cannot flex to accommodate demand. Del Negro and coauthors have decomposed the decline in r-star and attributed a substantial portion to precisely this convenience yield channel.

The consequences for the effective lower bound are profound. If safe asset scarcity depresses the entire term structure of risk-free rates, then conventional monetary policy runs out of room more frequently. Central banks find themselves at the ELB not because of temporary shocks but because of structural imbalances in the safe asset market that monetary policy alone cannot resolve.

This reframes several policy debates. Government debt is not simply a fiscal burden—it is also, from a functional standpoint, public liquidity provision. Reducing sovereign debt during periods of safe asset scarcity may impose real welfare costs by exacerbating the shortage of collateral and settlement instruments.

Takeaway

When safe assets are scarce, government debt becomes infrastructure. The optimal debt level cannot be determined by fiscal considerations alone—it must account for the economy's demand for public liquidity.

For monetary policy transmission, the convenience yield framework demands a substantial reconceptualization. The traditional New Keynesian model treats the policy rate as directly influencing the marginal decision-making of consumers and firms through consumption Euler equations. But if safe rates embed large and time-varying convenience premia, then movements in policy rates transmit imperfectly to the shadow rates that actually govern real decisions.

Quantitative easing programs illustrate this vividly. When central banks purchase long-duration Treasuries, they simultaneously operate through the portfolio balance channel emphasized by Vayanos and Vila, and through a scarcity channel that intensifies convenience yields on the remaining stock. These channels can have offsetting or reinforcing effects depending on the state of financial intermediation, complicating the calibration of asset purchase programs.

Sovereign debt management, historically viewed as an efficiency-and-cost problem for Treasury departments, becomes a first-order macroeconomic policy instrument. Greenwood, Hanson, and Stein have argued that the maturity structure of government debt should respond to fluctuations in the term premium and the demand for near-money assets. Issuing more short-term bills during periods of high near-money premia lowers financing costs and provides valuable liquidity services.

There is also a coordination dimension. When central banks and treasuries operate independently on the safe asset supply—one through QE purchases and reserve creation, the other through issuance decisions—their actions can inadvertently work at cross purposes. The consolidated government balance sheet perspective, championed by Greenwood, Hanson, Rudolph, and Summers, becomes indispensable.

Finally, the convenience yield lens reframes financial stability considerations. Regulations that mechanically increase demand for safe assets—well-intentioned as they are—can inadvertently deepen scarcity and amplify convenience premia, creating incentives for private safe asset production through securitization. This dynamic contributed to the shadow banking expansion that preceded the 2008 crisis and remains latent in the current system.

Takeaway

Monetary and fiscal authorities are joint producers of safe assets whether they acknowledge it or not. Policy coherence requires treating them as such.

The liquidity premium puzzle is not an anomaly to be explained away—it is a signal that our asset pricing frameworks have been incomplete. Safe assets carry a monetary dimension that traditional models systematically underweight, and recognizing this fact reshapes our understanding of interest rates, monetary transmission, and fiscal policy.

For central bankers, the implications are practical. Policy transmission depends on the state of the safe asset market, not merely on the announced policy rate. For treasury officials, debt management is macroeconomic stabilization by another name. For academic modelers, the challenge is to integrate convenience yields into DSGE frameworks that have traditionally assumed frictionless intermediation.

The puzzle also invites humility. Markets have long revealed something our theories struggled to accommodate: that liquidity itself is priced, that safety is scarce, and that the economy demands more from its financial infrastructure than pure discounted cash flow analysis can capture. Building better models begins with taking that revelation seriously.