For someone who has now traded bonds professionally for over 30 years, the recent behavior of the bond market has been nothing short of incredible. For those who think of the bond market as a shelter from equity market storms, the results have been mixed, to be charitable. On the one hand short-duration Treasurys have hung in quite well during the recent tumult. On the other hand, longer-duration bonds have lost a lot of value. This dismal performance of longer-duration bonds, both so far this year and in 2022 when the equity market fell almost 20%, has been incredibly disappointing to bond bulls.
To explore what is going on, let us take a fresh look at the fundamentals of the yield curve. The “atomic” risk-free asset is cash. Cash provides liquidity and store of nominal value. Just beyond cash is the Treasury bills complex, all the way from 4-week bills to 1-year bills. As we go from cash to T-bills we take a small amount of additional duration risk. Then of course is the world of Treasury notes and bonds. As we go from the shortest notes to bonds, the duration increases monotonically. The price volatility of a bond to leading order is simply its duration times yield volatility. Yes, there are other, higher order things such as convexity, but let’s keep it simple for this discussion and focus on duration. Duration is to the bond market that equity beta is to the equity market. Duration is the “risk-factor” that allows investors to quantify the risk-return tradeoff across the yield curve.
So, here is what seems to be going on:
1. Tariffs Are Inflationary
When tariffs and taxes on goods go up, prices go up. Prices going up is the definition of inflation. Since interest rates and yields are the sum of (1) expected short-term rates in the future, plus (2) expected inflation, plus (3) unexpected things called “risk-premium,” when inflation goes up yields go up. When yields go up, prices go down by roughly the yield change times the duration. The longer the bond, the longer the duration, hence the higher the impact of the yield change.
2. Bond Market Volatility Has Gone Up
Bond market volatility has risen substantially, exceeding the levels seen during the COVID-19 crash and within shooting distance of the peaks seen during the 2022 bond market meltdown. Longer duration bond prices are simply too volatile. When creating a portfolio that diversifies stocks by using bonds (such as “risk-parity”), higher stock volatility, which has also risen a lot recently, results in de-risking of both stocks and bonds. This follows from the fact that investors rightly try to contain total volatility within reasonable bounds, so increased volatility in both stocks and bonds mechanically results in de-risking by selling both assets. Higher bond-market volatility means a higher risk premium to hold bonds.
3. Demand/Supply Mismatch
There is a lingering fear of a buyer’s strike against U.S. assets, and not just the so-called bond vigilantes who express their displeasure with a country’s economic policy by dumping their bonds (see: Liz Truss): It is no secret that with the large tariffs being imposed against China, one of the largest buyers and holders of U.S. Treasuries, a natural tit-for-tat response is either an outright sale of Treasuries, or at a minimum reduced participation by them in bond auctions. Given the size of the budget deficit that needs to be financed each year, removal of a large buyer might result in a natural increase of yields. With a relatively small spread currently between short-term rates and long term yields, investors are wary of extending duration when faced with this specter.
My own belief is that the Treasury market is still very safe, and that if push comes to shove the Fed will: 1) buy unlimited Treasurys to cap yields; and 2) cut rates. But when and at what levels is hard to forecast. The path from here to there can be rocky and tumultuous to say the least. One other indicator in support of this thesis is that the spread between U.S. and German bonds has exploded almost 0.6% since “liberation day,” and gold has rallied significantly over the same period. Has the process of global exit from Treasurys already started into foreign bonds? We will have data on this over the next few weeks.
4. Levered Basis Trades
There is much talk about the “basis trade” blowup — again. Basis trades popular in the bond markets include the spread between the current (on-the-run) bond and the old bond, the basis trade between bond futures and the cheapest to deliver basket, the mortgage-treasury basis, the corporate-treasury basis, the cross-currency basis and many more. The one for today: the swap-spread basis. In a swap-spread basis trade a trader takes a view on the spread between Treasury bonds and the matched maturity interest rate swap. This “four-legged” trade works in the following way. Borrow money at some repo rate to purchase a Treasury bond. To hedge the duration of the Treasury bond, enter into a swap contract to “pay-fixed” on a swap. On the swap the pay-fixed coupon is exchanged for floating rate SOFR. T
hus the total trade becomes a spread between (1) the Treasury coupon and the swap coupon in the long end (2) the spread (in the other direction) between the repo rate on the purchased bond and the SOFR rate in the short end. One notable fact about this set-up is that the swap part is a derivative contract, and hence is levered for the maturity of the swap (e.g. 30 years). The Treasury purchase is financed with the repo financing, whose maturity is usually short and at the mercy of the bank providing the financing. Since most basis trades are run with leverage to magnify the spread, any change in the repo conditions can result in forced unwinds.
Earlier this year the spread between the 30-year Treasury and the 30-year SOFR swap went to a negative 0.8% spread. This means that the yield on the Treasury bond was higher than the swap rate by 0.8%. Since this four-legged trade hedges out the duration, this negative spread is perceived as a compensation for the leverage that a 30-year swap provides with no credit risk. To basis traders this negative 80 basis point spread (of course levered up 10 times would be 8%) seems like a relatively low risk trade. Except it isn’t — when someone exits and makes the spread even more negative, it leads to forced unwinds of all four legs, and in the process results in forced sale of Treasuries. After liberation day this spread went more negative by about another 0.2% sharply to about -1% presumably due to forced unwinds.
5. No Liquidity
The Treasury market is generally considered to be the underpinning of all asset classes, since it is used to discount the cash-flows of all assets. This requires that the bond market always be very liquid. It is the foundational building block of all assets. Unfortunately as the Fed is running off its balance sheet, it has created a liquidity vacuum and other participants are quite wary of stepping into the vacuum for fear of getting run over.
6. Little Correlation Benefit
The fact that long duration bonds have not proved any benefit over the last two equity market debacles, and indeed by some measure could be the reason for the equity market debacle, is beginning to gather some believers. If inflation is the long-term risk to the economy, then bonds (see above), should be lower in price, and via the discounting mechanism (recall everything is discounted by the yield curve), all assets, including equities, should be dragged down. Thus the correlation between stock prices and bond prices would be expected to be positive, not negative, in such a scenario. This means that there would be no benefit to holding bonds for diversification.
At the end of the day, bonds have one redeeming feature now. Unlike the period following the COVID-19 crash when global central banks went on an indiscriminate, price-insensitive buying spree of the bond market — they have some yield. But even this is an Achilles heel for longer duration bonds since shorter duration bonds are much safer. Take a 2-year Treasury which carries a yield of about 3.95% as of this writing. With a 2-year duration, yields have to rise more than 200 basis points for this bond to lose money. For a 30-year bond with yields of about 4.95% and a duration of about 16 years, a 0.3% rise in yields wipes out any gains. Given that these bonds are not working as expected, and yields are moving 0.3% a day, it is only rational for investors to shy away from longer duration bonds. For my money, T-Bills out to 5-year Treasurys is the place to be, with the 2-year note the sweet spot.
U.S. Treasury Duration is having a rough patch. At some point I hope to be able to suggest that it is a great investment to extend duration. We are not there yet.