A Back-Of-The-Napkin Analysis of Recessions and Bear Markets For Investors

Summary

• I went back to first-principles and analyzed 62 years of stock market recessions and bear markets to see what it might tell us about the future.
• Recessions happen every 6–7 years on average, and planning for large downturns is a necessary component of any investing strategy. Bear markets also happen every 6–7 years on average, and are highly correlated with recessions, typically with the start of the recession preceding the worst of the bear market.
• Investors may want to focus less on recessions and more on the corresponding bear market, mostly because recessions tend to get identified too late to be predictors, and bear markets have different durations which directly impact stock strategies.
• The recent OCT 2022 bear market wasn’t that severe. It is quite likely the “recession” already occurred, the bottom has already been reached, and recovery is in the works.
• This article covers what we know about recessions and bear markets, the relationship between them, whether they can be predicted, whether a recession is coming in 2023, and why this topic is relevant to investors.

The stock market goes through cycles of fear and greed, and frequently both at the same time. Credit: DALL-E

Introduction

I did a back-of-the-napkin analysis of the stock market from 1961 to present (62 years), in order to get a high-level direct analysis of the source data. Contrast this with the myriad of derivative interpretations and dogma prevalent in the mainstream media. Data, such as the dates of recessions, was sourced from wikipedia and overlaid on the chart. Note: time frames and percentages are roughly accurate due to the rapid analysis method used.

62 years of stock market history with bear markets and recessions tagged. Analysis and predictions dated: JAN-8-2023. Credit: self using TOS charts

How to read the chart: The pink bubbles on top are recessions. The blue bubbles along the bottom are bear markets (-20% SP500). The pink and blue lines show the duration and relative timing of recessions versus bear markets.

• I recommend you spend some time looking around the chart and draw your own conclusions now, prior to reading my analysis below.
• The following questions have been answered primarily based off of the above chart analysis, with additional context from other sources added when needed.

What is the definition of a ‘recession’ and a ‘bear market’?

There are multiple definitions of what is or is not a recession, and it can get complex.

Generic definition (simplistic, but at least precise, and faster):

Two consecutive quarters of negative gross domestic product (GDP).

National Bureau of Economic Research (NBER) definition (which is vague, highly subjective, and slow):

A significant decline in economic activity that is spread across the economy and that lasts more than a few months. While each of the three criteria — depth, diffusion, and duration — needs to be met individually to some degree, extreme conditions revealed by one criterion may partially offset weaker indications from another.

Bear markets are much easier to define:

Bear markets are often associated with declines in an overall market or index like the S&P 500, but individual securities or commodities can also be considered to be in a bear market if they experience a decline of 20% or more over a sustained period of time — typically two months or more.

In the chart, the pink lines show the NBER durations. Blue lines (bear markets) actually stretch from the previous record high (RH) point, through the 20%+ downturn, to the time where it returns to RH (so they are longer than just the period where it was -20%+). I do this because investors care about when the downturn is starting, whether it is one of the bigger ones, and when it has fully recovered back to previous highs — essentially how long is the overall downturn and what was its severity?

What do we know about recessions?

• Frequency: They happen every 6.9 yr on average. (The Balance says every 6 years, starting from 1945).
• Duration: They last 10.5 mo on average (from 2 — 18 mo). (Wikipedia says 10 mo, starting in 1945).
• Recessions might be less of a concern to investors than bear market indicators because they don’t directly map to stock prices (e.g. the recession might be officially over, but your stocks are still underwater.)

What do we know about bear markets?

• Frequency: They (also) happen every 6.9 yr on average.
• Duration: They last 2.5 yr on average (from .5 — 7 yr). Note: I am using a “Record High (RH) to Record High” definition for these durations (the period where it is -20%+ is shorter.)
• Relative Duration: Market recovery takes longer based on the amount it drops. A 10–20% SP500 drop (not a bear market) takes about 6–12 mo to fully recover back to RH. A 20–30% drop (bear market) takes about 6–18 mo to recover. A 30–40% drop takes about 3 years. A 40%+ drop takes 5–7 years. So the deeper it goes, the longer you’re going to have to wait it out.
• Severity: Bear market severity is mostly easily quantified in terms of drawdown (% drop from RH). By definition they are -20% or more. Here are some of the worst ones. In 1942 (start of WWII) the peak of the bull market drawdown hit around -76% (not shown in chart). In 1974, -47%. In 2003, -45%. In 2009, -54%. By comparison, the recent Oct 2022 downturn that has everyone panicking was only -25% (so far). There are other downturns that happen more frequently, that are less severe than the above examples. So it is theoretically possible that the current downturn will get worse than -25%, but it is far more likely that this is one of the common examples of a less-severe bear market.
• The vast majority of stocks purchased at a high-point don’t increase in value during bear markets. Typically for most ETFs and common stocks you have to wait for the entire market (SP500) to reach new record-highs in order to get new profits. Dollar-cost averaging (DCA) for example, advocates regular periodic buys and depends on new record high points coming later, as the overall market grows.

What is the relationship between a recession and a bear market?

• Recessions and bear markets happen just as frequently.
• Bear markets last much longer than recessions (partially due to my conservative definition of “historical high to next historical high” for charting bear markets.)
• Recessions and bear markets are highly correlated. Most of the time if there is a recession, there is also a bear market and vice-versa. But not always (e.g. in 1980 there was a recession with no bear market, and in 1987 there wasn’t a recession, but there was a bear market.)
• It is possible there is a causal relationship between the two variables. Logically if there is a “significant decline in economic activity that is spread across the economy” one would expect people to sell off stocks to pay other bills, and have less cash to invest in new stocks. Or perhaps the relationship is in the other direction. Regardless, if we presume there is a tight relationship, then cases where the market dropped 20%, but there was no official recession, might indicate that the NBER’s definition is too subjective, and that it actually occurred anyway.
• Recessions tend to start just before bear markets begin in earnest (see chart). Thus the formal start of a recession would theoretically be a good predictor of when the market is about to go south — except for the fact that it typically takes 234 days for the beginning of a recession to be retro-actively labeled as such, and over a year for the end to be identified. Even if it was an accurate predictor, the delayed reporting makes it useless for guiding investing behavior.
• If the market is getting close to 20% down, a recession has probably already begun (but won’t be reported for several hundred days).
• In 8 of 9 cases, the market returned back to record high (RH) prior to the next recession starting (1980–81 was a bit messy). So if the market hasn’t recovered yet, then a recession probably isn’t starting.

Can recessions be reliably predicted?

“Prediction” is a vague word and used in an imprecise manner by most people. We might do well to focus more on forming hypotheses about the underlying system, and empirically testing them over time, rather than constantly making unverifiable predictions. Predictions should be precise, testable, and have a clear duration. Most predictions in the business press use the words “may”, “might”, “could”, “is coming”, or offer long time frames (e.g. within the next several years). This shows the authors are uncertain, and don’t want to be pinned down, lest they be proven wrong.

Example of a useless prediction:

Wait for several years since the last recession (easy since it takes over a year for them to call the end of it). Then go find two variables that often cross over on a chart during volatile situations, and then predict a recession “in the next 2–3 years” when they invariably end up crossing. Then add a little fudge factor by saying “probably” or “about” in there somewhere, and be sure to omit an exact expiry date.

Since recessions happen every 6 years, but maybe every 4–8, you should have nearly 100% chance of having an “accurate prediction”. If you happen to get lucky, and it happens earlier than expected, you’ll likely be crowned the next hedge fund manager with “your finger on the pulse of the economy.”

One common “recession prediction” method touted by the media is the “inverted-yield curve”. As I have said before:

You may have heard of an “inverted yield curve” which predicts the onset of a recession. A “predictor” that indicates that something “sometimes happens” within 2–3 years AFTER the event, for an event that happens every 6–7 years anyhow, is not a mathematically useful prediction. A dead clock is right twice a day, and guessing heads or tails is right 50% of the time. In contrast to conventional wisdom, the stock market typically goes UP 20–30% after a yield-curve inversion — be skeptical of investing superstitions and numerology.

Based on the chart analysis, it is frequently the case that ~-18% in the SP500 means a recession has already started — regardless of whether the NBER has labeled it as such yet (or ever will due to technicalities and politics).

Negative GDP growth might be a good indicator of the start of a recession (or a bear market), because it happens fast enough to be used as a predictor to guide investing strategies (I will explore this in a later article).

Predicting when a formal recession will occur is probably not possible currently, due to the number of variables and the fact that they change unpredictably. However, predicting that we might be overdue for one (e.g. 6 yr with no recession), or finding telltale indicators when one has already started, may be much easier to do.

What about the “coming recession” of 2023?

In early 2023 (at the time of this writing) there is an enormous amount of fear that another recession is coming. In OCT 2022, 90% of CEOs said a recession was coming in 2023. Mass lay-offs of more than 125,000 people in late 2002 and early 2023 are being attributed to recession fears. All the experts must be correct, right? The CEOs with ivy-league MBAs wouldn’t simply conform to herd behavior, right?

What we know currently:

• Based on the chart, we know that after a bear market occurs, it almost always reached record-high (RH) before a new recession starts. That hasn’t happened yet.
• Back in Q1 and Q2 of 2022 GDP went negative, which by one definition, was the start of a recession. IF we presume that the start of a recession presages the fairly immediate start of a bear market (as inferred in the chart), then that bear market probably already bottomed out back in OCT 2022, and we are just waiting for the recovery now. Summary: the recession already occurred, the bear market following it already occurred, and we are headed back to RH before the next recession comes around.
• We just had a formal NBER-approved recession back in 2020. That was just over 2 years ago. Arguably, we had another one in 2022, just a few months ago. We are NOT overdue for a recession, which on average happens every 6 years.
• The current 2022 bear market has already lasted 13 months (using the conservative RH-RH measure as previously discussed). On average, bear markets last 2.5 yr (based on our graph analysis). However, IF the current bear market has already reached bottom, then at its peak it only reached -25% drawdown. Based on the graph, bear markets of that size tend to take about 6–18 mo to recover back to RH. If this holds true in this case, we should be looking at an SP500 that is back to RH by mid-late summer (to put a line in the sand, lets say the last day of August.) If the SP500 goes lower than -25% again, then the bear market is deepening and it will likely be a long time before it returns to RH. This scenario seems unlikely given the current mix of economic indicators and the rebound in the market in FEB 2023.
• My prediction is that the following will happen:
  1) The market will return to RH prior to the next recession.
  2) We probably won’t end up with a recession at all in 2023. The “recession” and associated downturn already happened in late 2022.
  3) In the unlikely event that we do, it will happen late in the year after the market recovers to RH and provides a convenience exit point for short-term investments.
  4) When the next recession rolls around, probably in 3–4 years, everyone will be claiming they “predicted” it in 2022, but by then they will have re-hired everyone from the layoffs they did to “prepare for it”. They will also have conveniently forgotten that it was supposed to occur in 2023.

Do recessions really matter when it comes to investing strategies?

It could be that investors are focusing too much on recessions. Whether the NBER thinks that a “significant decline in economic activity, including depth, diffusion, and duration” has occurred, is really less important to an investor than whether a bear market is about to happen or not (but they are related).

There are many different forms of investing, and perhaps there are some forms that are dramatically impacted by recessions or things that might accompany them. However, for stock investing, we really care about when a bear market is about to occur, and how long it will go on for until it is notching new record highs again.

Conclusion

This article goes back to first-principles to understand the relationship between recessions and bear markets for investors. It has developed a number of rules of thumb for how recessions and bear markets behave, which can be used to make better predictions of where we are now.

Tips

• Plan for unexpected bear markets in your long-term investing strategy. They happen roughly ever 6 years, and you could potentially get stuck in one for up to 7 years.
• Don’t get caught using leverage during the peak of a bear market. Easier said than done, but be sure you have enough padding to withstand a theoretical 76% drop in your portfolio value if a world war breaks out. Trust me — it’s easy to over-extend yourself.
• There are a lot of people who profit off of a scared market. You can get the highest returns (in some cases 5–600%) when stocks are artificially low. Avoid believing the fear-mongering that is a calculated strategy by certain market actors.
• Stop believing that recession timing can be predicted far in advance. Use your own logic and analysis of indicators to determine which way the market is heading.

Community Questions

• Do you think there is a recession coming?
• Do you think the market will go back to previous highs before the next recession?
• What do you think the best indicator that a bear market is in the early stages is?

Notes

• I do not hold any financial degrees or certifications. I am not a tax advisor. Your investment decisions are your own, and it is best to test strategies with small amounts of money first, preferably after extensive back-testing.