There are decades when nothing happens, and there are weeks when decades happen.Lenin
Yes, so in case you are wondering we ARE in the Lenin quotes phase of this crisis (Warren Buffet quotes was first week of March, Charlie Munger second week of March and Peter Lynch third week of March, since you asked).
Firstly, I sincerely hope you and your family are well. Clearly, thinking about the stock market is probably not top of everyone’s mind right now. Other things are far more important.
Having said that, if you have been thinking about the market recently, perhaps explaining what’s happened to clients or colleagues, then you might well have asked yourself one or more of the following questions:
When will the bear market come to an end?
How long will it take to recover?
Will the market bounce straight back?
These are some of the big questions we’re all asking right now.
The classic flippant answer is: if I knew when markets would bottom and how much they would bounce then I wouldn’t be here, I’d be … self-isolating on my private island. No one rings a bell at the bottom. True, perhaps, but really not that helpful. We can do better than that. History can provide useful context.
All bear markets are different of course. They have different causes and the the economy and financial system enter them in more or less resilient situations. Policy makers react in better or worse ways. They all have things in common though, human nature doesn’t tend to change that much and it is worth remembering that markets are just a collection of individuals (or algorithms, written by individuals – perhaps one key difference this time around) setting prices and making trades.
Human nature does come into it – panic, fear greed drive prices as much as fundamentals (more on that below). So past bear markets can provide some useful information. As Ray Dalio has said, cause-effect relationships remain similar through history as there are “timeless and universal mechanisms” underlying them.
Goldman Sachs wrote a useful piece looking at 27 bear markets (in the S&P 500) since 1835, they looked at time from the previous peak to the bear market trough, and the time from the trough to recovering the previous peak. They also looked at the size of the decline. While there was a good range of outcomes, the averages are interesting:
My first takeaways from this were:
- The average bear market takes roughly 2 years to go from the previous peak to the bottom. Two years! We’re barely a month into this one (although it feels like two years)
- The recovery always takes longer than the fall – roughly twice as long on average. There’s a big range though of course. The longest recoveries (from the 1835 crash and the Great Depression) took over 20 years, whereas recoveries from the early 90’s and early 80’s bear markets were just 3-4 months
- 30%+ falls are common in the stock market, Goldmans count 16 since 1835 (once every 11 years) and 27 falls of 20% (once every 7 years). Perhaps we should think of this as the “price of admission”. While every crash feels different, once a decade falls in value like this are just part and parcel of being a stock market investor (a feature not a bug, as some have said)
Goldman did something interesting though. They classified the bear markets by cause: cyclical (your standard sort of recession like 1990 or 1980-82), structural (a bubble pop like 2008) or event-driven (like 1987). Now if you buy into these classifications (and it’s a big “if”) then there’s reason to think this market fall could be more like Goldman’s “event driven” bear markets rather than the overall average. And they have different timelines:
- Shorter average time to bottom of 9 months
- Shorter than average time to recovery of 15 months vs 4-5 years on average for all bear markets
The other event-driven bear markets identified were 1987, and the lesser-known market crashes of 1966, 1961 and 1956 which all saw 20-30% declines. Either way, we should expect this to last for a while, with plenty of bear market rallies, as well as further falls.
Before anyone says it, yes, the analysis would be different if we looked at Japan 🙂
After the fall
The equity returns coming out of bear markets tend to be pretty good as well. PGIM have written a good piece analysing returns before and after volatility spikes defined by the VIX index. They look at 26 such spikes over the last 68 years (some of which overlap with the Goldman’s study). The heightened volatility lasted for an average of 9 months, with a range of 5-16 (again, we’re a month in to this one).
What they show is that during a post-spike period defined as “after the dust has settled” (a period of 20 months, after volatility has subsided, and only known in hindsight of course) equity markets typically returned 20-25%. There was a range, but on only 3 occasions were S&p 500 returns in this period negative.
Interestingly they also showed that previous dynamics relating to equity stocks and sectors that were “winners” pre-shock became disrupted, with pre-shock winners and losers doing equally well post-shock. They also showed that the classic Fama-French factors, particularly value tended to underperform the market in such post-spike recoveries.
The Fear Premium
Why have markets fallen so much? One way of looking at the overall level of the stock market is to see it as the expected (discounted) future value of the earnings of the companies within. DWS put out a good piece which looked at the question of how much markets should fall, if earnings behaved the same way they did in 2008 (they halved) and then returned to their pre-crash levels in about 2 years. This sort of fall in earnings explains up to about a 10% fall in the level of the market. So why are markets falling so much more than that?
Well, as is often the case the key lies in the discount rate being applied to these earnings. It turns out that the current price of a future stream of earnings is annoyingly sensitive to the discount rate that you use to bring future payments back to the present (just ask any pension fund actuary). At times like this investors demand a higher discount rate, which you can think of as the fear premium. for markets to return to similar levels to before the covid19 outbreak, we’d need that fear premium to subside as well as the underlying earnings to bounce back. And it’s important to remember that pre-virus, many equity markets (particularly the S&P 500) were priced “for perfection”.
Another way of looking at markets is to say that they attempt to weight a range of outcomes – some good, some bad. Obviously we now know what happened to corporate earnings following 2008 but at the time this was widely uncertain and the market reflected a wide range of possibilities. At the time of writing the range of outcomes here still seems pretty wide with some good and some very bad scenarios very much still in the picture, and it is this uncertainty that is driving things. No-one can be certain of anything at this point.
“V”, “U”, or “L”
That doesn’t stop economists putting out predictions of course and it has been common for economists to serve up and alphabet soup of predictions – whether recoveries look like a V, a U or an L. (I’ve also heard “check-mark” – V with a partial recovery leg and “swoosh” V with a longer recovery leg). The Economist ran a good piece noting some of the conditions that would need to be in place for a V-shaped recovery in the economy (and we should remember at this point that the stock market is NOT the economy):
- Minimal damage to infrastructure and institutions
- Trade networks not disrupted
- Macro-economic policy works well
Several economists are forecasting just such a recovery such as Schroders and Capital Economics. Others are suggesting that China offers some evidence that a V-shaped recovery is possible, but quite a lot needs to go right at this point for that to happen.
Others are making the point that wounds heal, but scars last. So the general level of risk-aversion that might apply after such a stark shock (individuals and firms keeping more cash and consuming, investing less, extending less credit) could hamper such a recovery, if behaviours change in the medium or long term. Ray Dalio goes further, theorising that we are at a seismic turning point, and with monetary policy having less and less scope to help, wealth inequalities magnified and with populism on the rise this could hasten a changing of the world order.
That most famous and over-used of John Templeton quotes is that the most dangerous words in investing are “this time it’s different”. Well, you could also argue the reverse – danger comes from an overreliance on past data and relationships “this time it’s the same” is perhaps just as dangerous. A safe bet is that this crisis will be different in some ways and the same in others, of course we’ll only know for sure after the fact.