Behold the Most Volatile Call for Bond Yields

(Bloomberg Opinion) -- Stay in one part of the market long enough and you're bound to know which strategists tend to be bullish and which ones seem permanently bearish. U.S. Treasuries are no exception. Just consider these two divergent views from the ranks of the Federal Reserve's primary dealers:

Steven Major at HSBC Holdings Plc is among those who have long believed in lower-for-longer U.S. interest rates. He and I talked in mid-2016, when Treasury yields hit all-time lows, about how structural forces such as an aging global population create a nearly insatiable demand for safe fixed-income assets. He predicts the benchmark 10-year yield will finish 2020 at around 1.5%, compared with a median estimate of 2% in a Bloomberg survey. That's bullish.

His polar opposite might just be Stephen Stanley, chief economist at Amherst Pierpont Securities. Dating to at least mid-2017, Stanley's prediction for the 10-year Treasury yield has always exceeded the median estimate. That hasn't changed for 2020 — he's the only analyst among the 49 surveyed to see the benchmark U.S. yield back at 3% at the end of next year. That's bearish.

And then there's John Dunham, who shatters the conventional labels. 

Dunham, managing partner at John Dunham and Associates, is more bearish on bonds in the coming months than anyone. He predicts 10-year yields will soar to 2.75% by the end of June and climb to 3.03% by the end of September. Even Stanley sees them merely gliding to 2.5% at mid-year and then 2.8% at the end of  the third quarter. In Dunham's scenario, current owners of 10-year notes would face a roughly 10% loss in 10 months.

If Dunham's right, though, those investors ought to hold on for dear life. By mid-2021, he suddenly becomes one of the biggest bond bulls on Wall Street, forecasting 10-year yields at 1.3% for the rest of that year. For those keeping track at home, that's an even lower forecast than Major's 1.5% estimate for year-end 2020.

This type of market swing, of course, is hardly unprecedented. In fact, the 10-year Treasury yield plunged from 3.25% in November 2018 to just 1.43% in early September as bond traders shifted from expecting more Fed interest-rate increases to rapidly pricing in easier monetary policy. That 182-basis-point range is right in line with the type of move that Dunham envisions. Just on Tuesday, long-term Treasury yields tumbled 10 basis points, the sharpest drop since August.

Still, few analysts (if any at all) actually come out and predict that sort of volatility. The tried-and-true playbook is to call for interest rates to rise gradually or fall from their current levels and then tweak those forecasts as the market moves. You just don't see a forecast slice through the median and average as drastically as Dunham's does.

So, what explains such a turn of events? To Dunham, it's fairly straightforward: Inflation will take off for a short period in 2020, followed by a recession after the U.S. presidential election (he hasn't predicted who will win it). He explained his view to me over the phone:

“I believe that the administration is going to just do everything it can to crank out money into the economy until the election, just to keep it going. And after the election, they're not going to ... That in many ways is driving our forecast for a recession happening right after the election. Usually the first quarter is terrible anyway, so that makes good sense that the first quarter would be the time we would tumble into recession.

When you start looking at business cost factors, the employment cost index has really been rising rapidly since the recession, we're seeing the dollar up a lot, that takes up prices. The only thing that's really been holding things down is commodity prices have been relatively flat. That's going to turn at some point.

We've been thinking — and I've been wrong about this, admittedly — that there will be decent inflation coming up for a while. We don't model the Federal Reserve as independent, we model it as a trailing factor. It follows interest rates. That's why we have them pushing up interest rates, up to that point that we hit recession.”

While Dunham doubts the Fed has any ability to stoke inflation, it's worth noting that just this week, the Financial Times published  an article that said the central bank was considering a rule that would allow price growth to run above its 2% target in a “make-up strategy” for years of undershooting its goal. That has interest-rate strategists like Mark Cabana at Bank of America Corp. thinking that current market-implied inflation rates are probably too low.

Dunham has called for a recession before. He said in mid-2015 that “we're now at the peak of the business cycle. And over the next year, year-and-a-half, the business cycle is going to start to turn back into recession.” While that didn't quite pan out, in that period here's what did happen: The U.S. stock market swooned, real gross domestic product nearly turned negative and 10-year Treasury yields fell to unprecedented lows.

His recession timeline also matches up with the historical signal given by the inverted U.S. yield curve. Three-month Treasury bills yielded more than 10-year notes for much of the period between late May and mid-October. That has usually indicated an economic downturn within 18 months or so. The curve from two to 10 years flipped to negative for a brief stretch in August. Cast those dates 18 months forward, and it's right around the turn of the calendar from 2020 to 2021.

Whether the world's biggest economy follows that traditional rule of thumb is anyone's guess. And, to be sure, the Fed has tried time and again to lift inflation only to see its preferred gauge remain stubbornly below 2% for almost all of the past decade. A lot will have to go right — and then wrong — for Dunham's forecast to play out.

At the same time, the likelihood that Treasury yields will hug the median in the coming year seems about equally as far-fetched. Last December, the median analyst forecast for where the 10-year yield would be at the end of 2019 was 3.32%. That's shaping up to be off by about 150 basis points. It was a closer call in 2018, thanks in large part to the end-of-year bond rally, but the December 2017 consensus still wound up missing by a quarter-point.

In other words, predicting the future is hard. Might as well forecast boldly.

I'm not including estimates that are a month or two from expiring, which all tend to converge to the prevailing yield level.

To contact the editor responsible for this story: Daniel Niemi at dniemi1@bloomberg.net

This column does not necessarily reflect the opinion of the editorial board or Bloomberg LP and its owners.

Brian Chappatta is a Bloomberg Opinion columnist covering debt markets. He previously covered bonds for Bloomberg News. He is also a CFA charterholder.

©2019 Bloomberg L.P.