The great bond rally

The ten-year US Treasury yield fell by more than 80 basis points in November 2023 from its October peak, ending near 4.17 per cent and pulling borrowing costs down around the world. It was one of the biggest bond rallies on record. What started as relief turned into a violent squeeze on positions.

From 5 per cent to relief

The starting point was extreme. In late October the ten-year yield briefly crossed 5 per cent for the first time since 2007, as investors demanded more compensation for holding long-dated debt from a government running large deficits. Mortgage rates rose to their highest in more than two decades. The turn came on 1 November, when the Treasury raised its auction sizes for long-dated bonds by less than feared and the Fed left rates on hold. A softer jobs report two days later and flat consumer prices for October did the rest.

Much of the October sell-off had been about the term premium, the extra yield investors demand for tying up money for ten years instead of rolling short-term bills. That premium had climbed as the deficit widened, as the Fed kept shrinking its bond holdings and as buyers worried about the flood of new supply. The Treasury's decision to lean more on bills and less on bonds went straight at that worry, which is why a dry announcement about auction sizes moved markets as much as a central-bank meeting.

Shorts forced to cover

Macro hedge funds and institutional investors had spent months building large short positions in duration. Cooler inflation data and softer words from the Fed caught them out. Once yields broke through technical levels, trend-following and momentum models kicked in and forced yet more buying to close shorts. The rally fed on itself, which shows that positioning moves markets about as much as the economic data do.

The process runs on autopilot. A trend-following fund sizes its bets by recent price moves, so once yields had fallen far enough to flip its signal, it had to buy back its shorts and then go long, whatever its managers thought of the economy. Discretionary funds with stop-losses were pushed the same way. Each wave of buying lowered yields further and tripped the next set of signals, which is why the move looked disorderly on some days even though the news behind it was modest.

Not all shorts are bets

The case against the squeeze story deserves a hearing. The large short positions visible in futures data belong partly to the basis trade, in which hedge funds hold cash Treasuries and sell futures against them to capture a small price gap. Those positions are hedged and do not need covering when prices rise. The rally also had a respectable fundamental base, since inflation did cool and markets moved to price rate cuts for next year. Even so, a move far quicker than the change in the data points to forced buying.

The basis trade carries its own risks, which have drawn warnings from the Bank for International Settlements and from American regulators this year. It is financed with short-term repo borrowing at high leverage, so a sudden rise in margin requirements or a disruption in funding markets could force funds to sell cash Treasuries in a hurry. In March 2020 exactly that happened, and the Fed had to buy bonds on a vast scale to restore order. November's rally posed no such threat, but the same positions would matter a great deal in a sell-off.

A precedent in reverse

The 1994 bond rout is the mirror image. The Fed's first rate rise in five years, a quarter-point move that February, hit a market full of leveraged long positions, and the rush to sell turned a policy adjustment into one of the worst years for bonds on record. November ran the same machine the other way. The lesson from 1994 was that positioning can exaggerate a move without changing where fundamentals eventually take prices, and that the overshoot tends to be corrected once the forced trades are done.

Mortgages and shares follow

The rally spread quickly beyond Treasuries. Freddie Mac's average rate on a thirty-year fixed mortgage, which peaked at 7.79 per cent in late October, had fallen well below 7.5 per cent by the end of November as lenders priced their loans off the lower ten-year yield. The broad investment-grade bond market had its best monthly return since the mid-1980s. Shares joined in, with the S&P 500 gaining almost 9 per cent in November, as lower discount rates lifted the value of future profits.

Lower yields abroad followed too. German and British government bond yields fell alongside Treasuries, partly on their own softer inflation data and partly because global investors arbitrage between the big sovereign markets. A borrower in Frankfurt or London therefore got some of the relief without any change in policy from the European Central Bank or the Bank of England, which is a reminder of how far American long rates set the price of money elsewhere.

What happens without the fuel

The winners from the rally are easy to name: mortgage borrowers, companies waiting to refinance, and banks sitting on unrealised losses on bond portfolios, which shrink as prices rise. From here the path depends on data and supply. The Fed meets on 12 and 13 December, the next quarterly refunding announcement comes early in the new year, and each monthly inflation print will test whether the cuts now priced are justified. If growth holds up and supply keeps climbing, yields have room to back up.

November's rally owed as much to bears being forced out as to faith in disinflation. With the shorts now largely covered, that fuel is probably spent, and further gains will have to come from the data. The more likely path over the next few months is a choppy range around 4 per cent for the ten-year yield, with a retreat towards 4.5 per cent if January's inflation figures run hot or the refunding brings bigger long-dated auctions.