February 15, 2013
Four Market Indicators Suggest That US Recession Risk Is Low
When recession risk is high, what signals do you expect from the markets? The empirical record and a deep library of research tell us to look for the following: a stock market that's posting negative year-over-year returns; a credit spread that's rising vs. year-earlier levels; a Treasury yield spread that's currently negative; and annual increases in oil prices. If all four danger signals were present, our cyclical goose would probably be cooked. But that's not the case at all. In fact, none of those warning conditions currently exist.
Do those four empirical facts guarantee that recession risk is low? No, but this market profile overall provides a strong clue for thinking that the economy, as we speak, isn't currently contracting. I'm the first to admit that you shouldn't look to market indicators alone for modeling the business cycle, which is why I routinely crunch numbers on a broad array of economic and financial data. (I'll have an update on the big-picture macro profile next week.) But we can and should start by looking at financial indicators.
Financial data, after all, is timely, offering a real-time profile of current conditions in terms of how the crowd is pricing risk. Even better, market prices are never revised. Conventional economic numbers, by contrast, arrive with a lag and are usually updated after the fact. We still must review economic reports, of course, but the information in macro data is worth a lot more when considered in context with market numbers.
On that note, let's consider the stock market (S&P 500), which is arguably the first choice for analyzing the market's perception of macro risk. History suggests that when the threat of recession is high, the equity market will price this hazard by posting negative returns on a year-over-year basis. Yes, it could be different this time, but saying so is just a guess, and one that has minimal support in the historical record. Meantime, let's recognize that six of the last seven recessions have been accompanied by negative returns, either in advance of a new recession's start date, or just after the downturn commenced. An accident of history? Perhaps, but you'll need a convincing model to explain why--models, to put it mildly, that are in short supply.
On a more practical note, it's true that the stock market has been known to fall on an annual basis without the onset of recession. That tells us to interpret negative annual comparisons cautiously in real time. These days, however, that's not an issue. The stock market's ahead vs. a year ago by a comfortable margin—roughly 10%. That's a fairly strong signal for rejecting the claim that the economy is currently in recession.
The credit yield spread, the Treasury yield curve, and oil prices offer corroborating evidence. In all cases, these markets don't appear to be pricing in the threat of a new recession. Yes, it's possible that all four market indicators are wrong this time. Heck, you can never say never in macro. But if we're trying to come up with a rough estimate of business cycle conditions in the here and now, these four metrics suggest that the odds are fairly low for arguing that the economy is in a recession today.
Our confidence is a bit higher once we look at a broad range of economic metrics in context with the market numbers. The big picture, in other words, continues to favor the case for growth. Modest growth, but growth nonetheless.
It could all change tomorrow, of course. But we don't have tomorrow's numbers yet. But pessimists aren't out of choices. They can still predict trouble next week, next month, next year. Before we do anything, however, let's start by recognizing what the published data tells us today vs. what forecasters think will happen tomorrow.
Posted by jp at February 15, 2013 6:18 AM
Treasury spread is 10yr less 3mo Tbill. Credit spread is Moody's BAA less AAA. Are the spreads likely to go negative these days? Probably not. Does that mean a recession is unlikely? That's the historical track record. Maybe it's different this time. Heck, it's always different this time in some respect. That's why looking at a handful of indicators, such as these four market metrics, should be considered in context with a wider set of data, which is why I do so regularly on these pages. That said, it's useful to look to the market numbers because they're continually updated and never revised. Looking at leading indicators that precede the leading indicators, sot to speak, is an interesting idea, but the usual caveats apply.
Posted by: JP at February 18, 2013 6:08 AM
How do you calculate the treasury yield spread ... 2s-10s? Would you really expect that might go negative in this environment? Same goes for the credit spread ... How do you calculate it and would you really suspect it might behave in the same way given what the Fed is doing?
Last but not least ... Given that the recession dating committee is often VERY slow in declaring the onset of a recession, and the stock market quite EARLY in doing so, would it not be more useful looking for indicators that precede a 20pct fall in stock prices?
Posted by: Tom M at February 16, 2013 8:36 AM