The Federal Reserve has a very poor track record of raising interest rates without pushing the economy into a recession
The Federal Reserve Board building on Constitution Avenue in Washington, U.S. Photo by REUTERS/Brendan McDermid/File Photo/File Photo With the Federal Reserve embarking on the fastest pace of tightening in nearly three decades there has been a growing discussion about its ability to engineer a “soft landing.” Fed Chair Powell even told Senators in a recent testimony that he believes achieving a “soft landing” is “more likely than not.” But the Fed has a very poor track record of raising rates without pushing the economy into a recession. In fact, over the past 14 hiking cycles, 11 have resulted in a recession.
Advertisement 2 This advertisement has not loaded yet, but your article continues below.
Recent research from a number of sources also cast doubt on Powell’s statement. As such, rather than buying into the Fed’s fantasy, inventors should exercise caution and shift away exposure from cyclical sectors into more defensive ones.
A new Centre for Economic Policy Research (CEPR) column and a National Bureau of Economic Reasearch (NBER) working paper co-authored by Larry Summers examined the plausibility of the Fed’s positive view on the economy amid rising interest rates. The researchers calculated the probability of a recession within the next 12 to 24 months based on historic relationships between economic activity, the CPI and the unemployment rate (the two most talked about economic variables recently) using data going back to the 1950s.
Advertisement 3 This advertisement has not loaded yet, but your article continues below.
They found that when inflation rises above five per cent in a given quarter, the odds of a recession over the next 24 months are above 60 per cent. And when the unemployment rate drops below four per cent, those recession odds are closer to 70 per cent.
However, it’s the combination of low unemployment and high inflation that significantly increase the probability of a recession. In fact, going back to 1955, there has never been a quarter with average inflation above four per cent and an unemployment rate below five per cent that was not followed by a recession within the next two years.
However, the probabilities above may be overstated given that business cycles have become more stable in recent decades, which reduces the probability of a recession. Additionally, the table above calculated recession odds based on historical data but didn’t give us any predictions about the future.
Advertisement 4 This advertisement has not loaded yet, but your article continues below.
With that in mind, the authors made use of a probit model to predict the probability of a future downturn controlling for current economic conditions. On top of that, to account for the possibility that recession probabilities have declined over time, they included a time trend in their model and other variables that show whether the economy is expanding or not. Their results still suggest a very high likelihood of recession in the coming years (up to 67 per cent in the next four
quarters and close to 100 per cent in the next 2 years).
The authors also note that there have been three instances in the past when the Fed did manage to engineer a soft landing — 1965, 1984 and 1994. These three periods have little to no resemblance to the current situation with the unemployment rate being much higher than today (there was more slack in the labour market in those periods) and interest rates also being well above the inflation rate. In their opinion, this only strengthens their resolve that we’re set for a much harder landing than before.
Advertisement 5 This advertisement has not loaded yet, but your article continues below.
More On This Topic David Rosenberg: Declining equity risk premium turning TINA argument around David Rosenberg: How investors can navigate the global water crisis David Rosenberg: Three investing themes beyond energy in these uncertain geopolitical times Additionally, these alleged “soft landings” are also event-driven. In the mid-1960s, the stimulus from the Great Society fiscal plan acted as a buoy for the economy. In the mid-1980s, oil prices plunged by nearly 60 per cent, which acted as a de facto tax cut for the consumer and in the mid-1990s, Netscape’s IPO triggered the dot-com boom, which in turn led to unprecedented wealth creation for five years. And let’s not forget that these periods also did not involve an ongoing pandemic, which has resulted in a major supply shock — something that is completely out of the Fed’s control.
This advertisement has not loaded yet, but your article continues below.
Article content What about other models? The literature on recession probability models is dense. We can use lower-frequency models like the one described above with inflation and unemployment as inputs. Or we can use high- frequency data from the yield curve (which is also highly reliable). In fact, our internal model that utilizes the 2-year/10-year curve is currently giving us close to 40 per cent odds of a recession over the next 12 months. This probability has declined slightly due to the recent steepening of the curve; but it remains at the highest level since early 2020.
Even if we use these models, the problem is that we rarely know when we have entered a recession as it can take the NBER close to a full year to call one. Former Bank of England Monetary Policy Committee member David Blanchflower, in a series of NBER working papers, suggested using consumer sentiment surveys as a proxy for recessions because they are timely and not subject to revisions. To identify a recession, he proposed a rule of thumb — a 10-plus-points decline from the peak in the University of Michigan (UMich) or Conference Board (CB) consumer sentiment indices coupled with a slowdown in the labour market (a decline in employment and a 0.3 percentage point rise in the unemployment rate in successive months). Both the UMich and CB indices have already declined by at least 20 points from the peak whereas the labour market has continued to strengthen. But the fall in the S&P 500’s employment services subsector could be indicating a slowdown in hiring, thus pointing to a turning point in the labour market. So, the next few months may provide more evidence of a downturn in the making.
This advertisement has not loaded yet, but your article continues below.
Article content Bottom line: Regardless of what model we look at — high-frequency, low-frequency, or sentiment indices — they all seem to indicate that a 12th recession following a Fed rate hiking cycle is more likely than not. This contradicts what Chairman Powell and many other market commentators are saying, but the truth is that the only way the Fed can tackle this sort of inflation (which is largely due to supply factors it cannot control) is by destroying demand and generating significant economic slack. For investors, this would translate into exercising caution and shifting away from exposure from cyclical sectors to more defensive areas of the market such as consumer staples, healthcare, and utilities.
David Rosenberg is founder of independent research firm Rosenberg Research & Associates Inc. You can sign up for a free, one-month trial on Rosenberg’s website.
Financial Post Top Stories Sign up to receive the daily top stories from the Financial Post, a division of Postmedia Network Inc.
By clicking on the sign up button you consent to receive the above newsletter from Postmedia Network Inc. You may unsubscribe any time by clicking on the unsubscribe link at the bottom of our emails. Postmedia Network Inc. | 365 Bloor Street East, Toronto, Ontario, M4W 3L4 | 416-383-2300