This article is an on-site version of our Unhedged newsletter. Sign up here to get the newsletter sent straight to your inbox every weekday
Welcome back, and thanks for all the feedback on yesterday’s inaugural edition of Unhedged. Sign up here to get this newsletter sent straight to your inbox. And keep letting me know what you think: robert.armstrong.
What to expect when you are expecting poor long-term returns
Yesterday I wrote that equity investors face a dilemma: very high valuations suggest poor long-term returns, but being out of the market risks missing out on the great short-term returns that often characterise the final leg of a bull market.
I am appalled to discover that not everyone agrees with me. A perceptive reader, Emeraldo560, did not much like the chart I used to make my point:
. . .[your] data is only from 1988, which given [that your] chart shows monthly overlapping 10 year returns, isn’t that much. Those points at the very right of the chart are mostly dotcom bubble and pre 2008/9 GFC (great financial crisis)
My friend Chris Rossbach, who manages money at J. Stern & Co. texted me this:
. . . [your] argument about high valuations . . . is wrong. Do a chart of dispersion of valuations. Yes average P/E is at the higher end (but still not that high if you capitalise it using current and likely future interest rates). BUT there are lots of companies that are at low multiples in absolute terms and relative to their earnings and cash flow growth.
Here again is the chart Emeraldo objected to, which I used to illustrate the valuations/returns relationship yesterday. It shows 10 year returns following various monthly valuation observations:
Maybe Emeraldo will like the one below better. It’s from Strategas, goes back to 1950, and it used the Shiller P/E, which uses long-term average earnings in the denominator, to smooth out cycles and shocks (I added the little arrows). The Schiller P/E is now at 37:
There are still not as many individual observations on the right side as you’d like in order to be really confident in what will happen next. You are still looking mostly at just a few, mostly overlapping 10-year periods. Still, there are more. And remember that this statistical analysis rests on an almost tautologically true hypothesis: that when you pay a much higher price for future cash flows, those future cash flows are likely to be smaller, relative to that high price, than if you had paid a lower price. Duh.
Around the current valuation levels, there have been some cases where returns have been OK. So maybe we’re fine. But there are also zero instances of juicy long-term returns near these valuation levels, and if valuations rise from here, all the juice is squeezed out.
On to Chris’s objections. He thinks that the answer to the investor’s dilemma is — to simplify — to pick stocks, rather than own the whole market. It’s an important point. A lot of managers (not Chris, by the way) think that the move now is to pick value stocks, which have underperformed until recently.
The previous two eras of bonkers high valuations were 1999 and 2007. Would have shifting to value helped then? Here is how growth and value indices did from 1999 to 2009, via Bloomberg:
Owning value helped a lot after the end of the dotcom bubble. You still had to eat a lot of bark over those 10 years, but you ended positive, even after the GFC. Now 2007-2017:
This is a very encouraging chart. Stocks did fine in the decade after the 2007 top. Drinks on me!
Value underperformed rather than being a safe haven, meaning that the value/growth call is a lot harder than “market top = buy value”, damn it.
Valuations were much lower at the 2007 top than they are now, the Shiller P/E was about 27, meaning those returns were right where that Strategas chart would have predicted.
You had to get through a 50 per cent drawdown in order to earn that adequate long-term return, during which both your clients and your spouse left you, and your dog started to have his doubts as well. Ready to gut that out again?
The reason you got those good returns is, in large part, because of massive Fed and government intervention after the housing bubble burst. Let’s suppose we get, in the medium term, the third major market-led economic crackup in the last 25 years. Are we going to get the same sort of government response, and is it going to work as well? In other words, are we in a permanent bubble-crash-bailout cycle, and is that — fine?
Chris’s second objection, that valuations are not that high if you index them to interest rates, touches on a huge debate. More on that at some point. But for the rest of the week I’ll stay away from valuations.
Volatility is a volatile way to make a living: a parable
Look at this SEC settlement, which landed Monday. Here’s a summary:
The CBOE operates an index of expected market volatility, called VIX, based on investors’ purchases of put and call options. S&P Dow Jones Indices, in turn, runs an index on top of the CBOE one, which uses the CBOE data to replicate the return from being long short-term volatility futures contracts. Credit Suisse, in turn, offered a sort of bond built on top of the S&P index, but upside down, so you could get the return on being short those volatility futures. People made good money on these Credit Suisse “exchange traded notes”, called XIVs, in the late teens. Volatility kept falling, and it was fun.
Then one famous day in February 2018, volatility went way, way up. And in the late afternoon of that day, for an hour or so, the S&P index stopped tracking the spike. Instead, it stayed flat. A “quality control” feature in the index made it simply reproduce the index’s last value whenever a much higher or lower reading came in (the public had not been informed of the existence of this feature). The people running the index could have overridden this feature, but they didn’t. So the XIV didn’t move during that period, either, when it should have been falling through the floor. When the XIV does fall through the floor, Credit Suisse is allowed to call — that is, cancel — the XIV notes outright.
The SEC sums up:
during the 4:00 PM hour and until 5:09 PM, when the closing indicative value of XIV was published, investors did not know that they had been purchasing and/or holding a product that had an economic value that was substantially less than what [S&P] had publicly reported and that was at risk of being [called] by its issuer
The next day, Credit Suisse did call in the XIV notes and shut down the product permanently. The fun was over. Now the SEC has settled with S&P indices, which has not admitted or denied anything, for $9m.
I’m not sure how important this weird little tale from the world of finance is. I’m not sure how many people were hurt by S&P’s price-reporting screw up, or were hurt by being dopes who think that volatility always goes lower. The story does, however, illustrates several eternal truths, to wit:
Complexity is to be viewed with suspicion.
Markets follow trends. Following those trends feels like making free money. It is not. Trends always change suddenly and with little warning. This is Benoit Mandelbrot’s great insight. Read his wonderful book.
(closely related to 2) The small print doesn’t matter until it does. The financial instrument you are using is fine, until it is tested by volatility, and all is gained or lost. Read the contract.
Credit Suisse struggles in investment banking.
One good read
I keep thinking about last week’s excellent piece in the New York Times about Bank of America honcho Thomas Montag. It made me think about how hard it is to have strong leadership in a Wall Street organisation without it becoming, to a degree, a personality cult. Legions of management consultants will tell you how it can be done. But I believe it’s very hard to avoid.