I’d wager that you could go on almost any news website today and find at least one article about the next recession. And most financial news sites have multiple articles exploring different angles on the recession. What is a recession? Will there be a recession? Which economic indicators indicate a coming recession or not? How long and severe will it be? How to prepare for a recession. How to invest in a recession. How will it affect Wall Street? How will it affect Main Street? I could go on.
So, with some trepidation, I’m adding to the blare of recession babble in today’s post. That’s because I still have a couple of questions that haven’t received much attention, as far as I’ve seen. Those questions are:
- What effect do recessions typically have on long-term investors?
- Why are so many expecting a recession, when considerable economic data still belie that expectation?
After some research, here are my answers.
Question 1: Recessions and Long-Term Investors
When it comes to the question of potential damage to your investment portfolio, most articles I’ve recently read focus on worst-case losses for stocks and bonds from past recessions. For stocks, here’s a good example, which I adapted from some RBC Capital Markets research as summarized by Sam Ro at TKer.
Like all stock data presented in today’s post, the surrogate for U.S. stocks in this graph is the S&P 500. The dates at the bottom of the graph are when each recession ended. The average peak-to-trough stock loss across these recessions was -32%. Ouch! So, long-term stock investors should expect some pain with the next recession.
But let’s take a closer look at these scary data. They represent the decline in stocks from the peak proceeding (or during) the recession to the subsequent trough around the same recession. This measures the loss for an investor who bought in right at the absolute top of the peak and sold right at the absolute bottom of the trough. While such people undoubtedly exist during each recession, the chances that any one investor would be so unlucky are extremely small. And any deviation from this poorest-possible market timing necessarily results in a less severe outcome for the investor.
Since 1928, recessions have ranged in duration from two months (COVID recession of 2020) to five years (the Great Depression). Because mindful investors are long-term investors, your investing timeframe is hopefully much longer than a few months. So, what if we instead examine the impacts on returns for whole years throughout the recession for each such event since 1928? Limiting the analysis to whole-year return data gives us a more realistic sense of what an investor with a 1- to 5-year timeframe might experience in the next recession.
The blue bars in the graph below show the same peak-to-trough declines as the previous graph. The orange bars with data labels show the cumulative nominal returns for whole years covering the duration of each past recession.
Simply expanding the investing timeframe to as little as one year paints a different picture of recessions. For example, consider the recession of 1953, which started in July 1953 and ended in May 1954, a period of less than one year. The peak-to-trough decline was -15%, but the total cumulative nominal return for someone who invested throughout 1953 and 1954 was an amazing 51%. Investing in stocks during past recessions resulted in positive nominal returns in 11 out of 15 cases with a median cumulative return of 18%, and after inflation adjustment, the median was still a decent 15%. I should note that the Great Depression (ending Mar-33) is a clear outlier, but few people are expecting the next recession to rival the worst one in history.
We can look at these recession periods in other ways. For example, here’s a graph that looks at annualized inflation-adjusted (real) returns for whole years covering recessions for both stocks and bonds (10-Year Treasury Bond).
Annualizing and, more importantly, inflation-adjusting the data suggests a slightly less rosy history for both stocks and bonds during recessions, particularly during the 1970s when inflation was very high. But even here both stocks and bonds produced positive annualized real returns during the majority of recessions. And the median for annualized real stock returns during recessions was 7.4%, and for bonds, it was 4%. Not great, but not awful.
And here’s a graph showing real cumulative returns for stocks and bonds during recessions, which shows a similar pattern as annualized returns but gives a better sense of the total damage (or lack thereof) across recessions of varying lengths.
Again, only 5 of the 15 recessions caused substantial losses for stock investors. And only 2 of 15 recessions caused substantial bond losses.
Before Recessions
It’s often assumed that stock markets trade on future expectations, and the same can be said to some extent for bond markets too. That is, market participants are thought to be buying and selling based on how healthy they think the economy will be in the future, among other considerations. The markets are sometimes totally surprised by unpredictable events. The most recent example is the abrupt crash of March 2020, when people suddenly realized the likely magnitude of COVID’s economic impacts. Nonetheless, before the onset of most recessions, there have been either clear or mixed signals that the economy might be worsening.
So, it’s reasonable to examine the whole year of returns before the official start of past recessions. Perhaps that might show that most of the returns damage occurred in advance.
The median real annual return for the years before recessions was 10% for stocks and 0.3% for bonds. And if we include one year before each recession using the same analysis above, we get this graph for real cumulative returns around the time of recessions.
The years before each recession only infrequently involved substantial losses. So, adding a prior year to the analysis doesn’t appreciably alter the picture of returns during recessions. In fact, in quite a few cases, adding the year before the recession improved the long-term performance of stocks or bonds. Buying stocks a year before the Great Depression and the recessions of the 1940s and 50s boosted asset performance. And the 2001 recession and early 1970s recessions are interesting counter-examples.
This suggests that either the markets are a pretty poor predictor of future recessions or they tend to look past the recession and further into the future. I could see lively debate on this question, but I think it’s largely unprovable either way.
After Recessions
As encouraging as all these statistics are, they’re still pretty pessimistic for a true long-term investor. After all, the longest period examined in any of the above stats is 6 years, whereas historical return data suggest that a holding period of at least 10 years is the safest bet for stocks. So, what if we look at even longer periods after recessions?
Fortunately, Ben Carlson at A Wealth of Common Sense has already done this heavy lifting for me. He calculated the cumulative returns for periods of 1, 3, 5, and 10 years after a recession has ended as shown in this graph that I adapted from his results table.
If you’re a mindful investor who buys and holds for the long-term, you have little to fear from recessions.
Question 2: Skepticism About The Next Recession
Because of the cyclical nature of the economy and markets, it’s not a question of if, but when the next recession will occur. And the media are brimming with predictions of a recession occurring before the year ends.
But all this imminent recession talk seems at odds with my recent experiences. For one example, when I picked up someone at the airport last week, the terminal was a giant rugby scrum of travelers, even though airline prices are at all-time highs, as this FRED graph shows.
Tracking data indicate that travel spending has almost entirely recovered to pre-pandemic levels, which says to me that consumers still have plenty of money to throw at one of the most discretionary of expenses.
Admittedly, my observation is anecdotal at best. But I’m not the only one questioning the recession narrative. Here are a few broader observations that I picked up from an article by Robert Fuller. One, the June retail sales report came in with a better than expected 1% increase for the month. Here’s a graph from Sam Ro at TKer.
Despite higher prices, people are still spending robustly across a range of products.
Two, consumer sentiment is still miserable, but there was a slight uptick in June along with easing inflation expectations as shown in this Bloomberg graph.
Three, supply chain problems that are exacerbating inflation seem to be easing, which means that the Fed may feel less pressure to aggressively raise interest rates over the next few months.
And here are a few other observations from some other sources. Four, recent bank earnings were way better than expected including clear indications of robust credit and debit card spending (see above), elevated credit card payment rates, low levels of credit losses, and strong demand for loans (except home loans).
Five, people seem comfortable with spending money because they have good jobs. This graph also from Sam Ro shows that employers are finding it harder to fill positions than at any time in recent history.
Six, surging wage growth is creating extra money to spend. Again, here’s a graph from Sam Ro.
On the other hand, this looks like exactly the kind of price-wage spiral that the Fed feels compelled to squash with higher interest rates, which could cause a recession. And there are some signs that labor markets are cooling, but that doesn’t necessarily mean a recession is imminent.
Compounding Uncertainties
If nothing else, all these conflicting economic data indicate the difficulty of predicting recessions. And when you try to extrapolate from the economy to the performance of stocks and bonds the uncertainties compound. For example, let’s pretend that we magically know for sure that a recession will start next month. And we know for sure exactly how much the Gross Domestic Product (GDP) will decline and how much unemployment will spike. Even with all that magical information available, history suggests almost no ability to predict how stocks and bonds will perform.
These cross plots of GDP declines and unemployment levels as compared to stock and bond returns from past recessions show the unpredictability of markets even during recessions. The red dots in these graphs indicate the recession periods when returns were negative.
So, even if the recession predictions come true tomorrow, we are still several steps away from knowing whether our portfolios will suffer any more than they already have.
Conclusions
I’ve focused today on positive economic news. However, I want to clearly state that I’m not predicting that we will avoid a recession. I often say that predicting the future in any meaningful way is impossible, so one shouldn’t try. As a forward-looking statement, the most I will say is that the economic picture looks cloudy to me. If I was forced to make bet, I’d assume that the odds of a recession in 2022 or 2023 are something like 50/50.
The more important take-home message is that if you’re worried about the damage a recession may cause to your portfolio, history indicates that worry is a waste of time, like most worries. No one knows for sure whether a recession is imminent. And even if it happens, there’s certainly no guarantee that it will cause mayhem to your portfolio. The real danger is letting your worries lead you to impulsive decisions when those temporary market dips inevitably occur.
