Reflation on the Mind


If you’d like to receive our market commentary by email every month, please click here
.

February saw the swiftest repricing of interest rate risk since 2016, with the 10-year yield rising up to 1.50% from a low of 0.50% in August 2020. The “savage” sell-off of bonds only saw relief on the last trading day of the month. The move was accompanied by impaired market liquidity and a so-called “failed” 7-year auction (with the largest “tail,” or auction concession, on record: 4.4 bp). The 10-year repo rate fell as low as -4.25%, which means holders of 10-year Treasuries were paid 4.25% per annum to borrow cash against the notes as collateral. Commentators have speculated that traders were betting against the 10-year en masse in anticipation of a stronger-than-expected recovery. Anecdotal evidence suggests that the interest rate moves were exacerbated by Japanese investors, which sold $34 billion in foreign bonds ahead of their March fiscal year-end rebalancing.

Given the complexity and fast-moving pace of the current market environment, this month we analyze February’s violent rates market sell-off and what it may mean.

Full Post

Reflation on the Mind

February saw the swiftest repricing of interest rate risk since 2016, with the 10-year yield rising up to 1.50% from a low of 0.50% in August 2020. The “savage” sell-off of bonds only saw relief on the last trading day of the month. The move was accompanied by impaired market liquidity and a so-called “failed” 7-year auction (with the largest “tail,” or auction concession, on record: 4.4 bp). The 10-year repo rate fell as low as -4.25%, which means holders of 10-year Treasuries were paid 4.25% per annum to borrow cash against the notes as collateral. Commentators have speculated that traders were betting against the 10-year en masse in anticipation of a stronger-than-expected recovery. Anecdotal evidence suggests that the interest rate moves were exacerbated by Japanese investors, which sold $34 billion in foreign bonds ahead of their March fiscal year-end rebalancing.

The narrative in the real economy has been consistent with the recovery trade. The jobs market continues to outperform expectations, with the February non-farm payrolls release showing gains of 379k vs consensus estimates of 200k and January estimates being revised upwards. The second stimulus package from December continues to make its way to households while a third and larger $1.9 trillion package (the American Rescue Plan) was passed by the Senate, with the House expected to approve the Senate modifications to the bill this week. Last month also saw the approval of the Johnson and Johnson vaccine, which subsequently began being distributed. As discussed in our December commentary, we expect equity market returns to remain choppy as sector rotations flow through the market and the 5s30s yield-curve transitions towards a steady state in the 175-220 bp range.

Given the complexity and fast-moving pace of the current market environment, this month we analyze February’s violent rates market sell-off and what it may mean.

The Cost of Time

In the rates markets there is the concept of a “term premium” which is simply the amount by with an investor is compensated for locking in a 10-year rate rather than investing overnight every day, over the same period. Theoretically, the 10-year Treasury interest rate at any given time should be exactly equal to the expected average overnight interest rate compounded over ten years. If it is higher, then an investor can borrow at at overnight rates, every day for ten years, to finance the purchase of 10-year notes and make the cumulative spread over the ten year period. This spread is the term premium.

In practice, the term premium is typically positive. This is because investors who are, for example, buying 10-year Treasuries want to be paid for assuming the risks of locking in a 10-year rate fixed-rate. By locking in a 10-year fixed-rate rather than invest in a floating overnight rate, such investor would not benefit from any unexpected increases in the Fed’s overnight policy rates.

The term premium is not something that is readily observable, but there is a model created by the Federal Reserve Bank of New York that estimates market-implied term premiums. The New York Fed model shows an average 10-year term premium of 1.50% during 1961 to the present. This means that on average, over the past sixty years, investors were compensated 1.50% per annum for locking in a 10-year fixed-rate rather than just lending short term in a floating rate (such as by buying Treasury bills or money market funds) every day for ten years. The plot below shows results of the New York Fed’s model:


Source. Federal Reserve Bank of New York

As expected, the late 70s/early 80s era of stagflation saw the highest term premiums of 3.0 to 4.0% per annum. Because inflation risks were firmly to the upside, investors demanded a significant term premium for bearing the risk of committing to a fixed-rate and foregoing the opportunity to benefit from any unexpected interest rate hikes to the overnight rate.

But the past few years saw an unusual phenomenon: the term premium became negative. This means investors were willing to pay to lock in 10-year rates– -0.5% to -1.0% per annum–rather than be compensated for assuming the risk of locking in a fixed-rate. The market essentially said that the balance of interest rate risk was firmly to the downside, perhaps because investors were concerned about deflation or negative rates and would prefer to lock in a fixed rate rather than be exposed to the specter of negative interest rate policy.



Source. Federal Reserve Bank of New York.

Turning to the current environment, we can refer to the New York Fed’s term premium model, which decomposes prevailing market interest rates into the (i) term premium and (ii) the average expected geometric mean overnight floating rate at any maturity. Below we do so and plot the expected average overnight floating rate over ten years (in black) and the actual nominal 10-year yield (in red). (The difference is the term premium.)


Source. Orthogonal, Federal Reserve Bank of New York.

The nominal 10-year yield (in red) has been on an upward trajectory since August 2020. However, over the same timeframe, the average expected overnight floating interest rate over ten years (in black) has trended lower. What the plot shows is the reversion of term premia from negative values to positive values (indeed, this reversion can also be seen at the tail end of the first chart above). Critically, this plot suggests the recent rise in yield was driven by this normalization of term premia, not an increase in expected rates. This means investors are less concerned about negative rates or deflation, and are once more seeing inflation as a risk worth being paid to bear. It is also clear the market is not expecting a rise in policy rates. Indeed, the Fed has repeatedly stated they are not thinking of raising rates in the foreseeable future and are willing to tolerate higher-than-target inflation. The rates market’s attitude towards the balance of inflation risk is becoming consistent with that messaging, and perhaps with the expectation of a stronger than expected recovery.

This is a welcome development for the Federal Reserve. For years, they have been trying to nudge inflation expectations up toward their target 2% level. Perhaps this shift in the rates market does not only reflect recovery expectations, but will be self-fulfilling, awakening the “animal spirits” in U.S. corporations and households to invest and purchase real assets and stimulate the economy as it emerges from nationwide lockdown. In any event, it seems unlikely that economic growth will be checked by a shift in Fed policy, or that the recent rise in rates is a sign of tightening financial conditions.