I would be lying if I said that I like down markets more than up markets, but I have learned to accept the fact that markets that go up will come down, and that when they do so quickly, you have the makings of a crisis. I find myself getting more popular during these periods, as acquaintances, friends and relatives that I have not heard from in years seem to find me. They are invariably disappointed by my inability to forecast the future and my unwillingness to tell them what to do next, and I am sure that I move several notches down the Guru scale as a consequence, a development that I welcome. To save myself some repetitions of this already tedious sequence, I think it is best that I pull out my crisis survival journal/manual, a work in progress that I started in the 1980s and that I revisit and rewrite each time markets go into a tailspin. It is more journal than manual, more personal than general, and more about me than it is about markets. So, read on at your own risk!
The Price of Risk
For me, the first casualty in a crisis is perspective, as I find myself getting whipsawed with news stories about financial markets, each more urgent and demanding of attention than the previous one. The second casualty is common sense, as my brain shuts down and my primitive impulses take over. Consequently, I find it useful to step back and look at the big picture, hoping to see patterns that help me make sense of the drivers of market chaos.
It is my view that the key number in understanding any market crisis is the price of risk. In a market crisis, the price of risk increases abruptly, causing the value of all risky assets to drop, with that drop being greater for riskier assets. While the conventional wisdom, prior to 2008, was that the price of risk in mature markets is stable and does not change much over short periods, the last quarter of 2008 changed (or should have changed) that view. I started tracking the price of risk in different markets (equity, bond, real estate) on a monthly basis in September 2008, a practice that I have continued through the present. Getting an a forward-looking, dynamic price of risk in the bond market is simple, since it takes the form of default spreads on bonds, and FRED (the immensely useful Federal Reserve Database) has the market interest rates on a Baa rated (Moody's) bonds going back to 1919, with data available in annual, monthly or daily increments. That default spread is computed by taking the difference between this market interest rate and the US treasury bond rate on the same date. Getting a forward-looking, dynamic price of risk in the equity markets is more complicated, since the expected cash flows are uncertain (unlike coupons on bonds) and equities don't have a specific maturity date, but I have argued that it can be done, though some may take issue with my approach. Starting with the cumulative cash flow that would have generated by investing in stocks in the most recent twelve months, I estimate expected cash flows (using analysts' top down estimates of earnings growth) and compute the rate of return that is embedded in the current level of the index. That internal rate of return is the expected return on stocks and when the US treasury bond rate is netted out, it yields an implied equity risk premium. The January 2015 equity risk premium is summarized below:
That premium had not moved much for most of this year, with a low of 5.67% on March 1, and a high of 6.01% in early February, and the ERP at the start of August was 5.90%, close to the start-of-the-year number. Given the market turmoil in the last weeks, I decided to go back and compute the implied equity risk premium each day, starting on August 1.
Note that not much changes until August 17, and that almost all of the movement have been in the days between August 17 and August 245 During those seven trading days, the S&P 500 dropped by more than 11% and if you keep cash flows fixed, the expected return (IRR) for stocks increased by 0.68%. During the same period, the US treasury bond rate dropped by 0.06%, playing its usual "flight to safety" role, and the implied equity risk premium (ERP) jumped by 0.74% to 6.56%.
I did use the trailing 12-month cash flows (from buybacks and dividends) as my base year number, in computing these equity risk premiums, and there is a reasonable argument to be made that these cash flows are too high to sustain, partly because earnings are at historic highs and partly because companies are returning more of that cash than ever before. To counter this problem, I assumed that earnings would drop back to a level that reflects the average earnings over the last 10 years, adjusted for inflation (i.e., the denominator in the Shiller CAPE model) and that the payout would revert back to the average payout over the last decade. That results in lower equity risk premiums, but the last few days have pushed that premium up by 0.53% as well.
My computed increases in ERP, using both trailing and normalized earnings, overstate the true change, because the cash flows and growth were left at what they were at the start of August, a patently unrealistic assumption, since this is also an economic crisis, and any slowing of growth in China will make itself felt on the earnings, cash flows and growth at US companies. That effect will take a while to show up, as corporate earnings, buyback plans and analyst growth estimates are adjusted in the months to come, and I am sure that some of the market drop was caused by changes in fundamentals. The argument that a large portion of the drop comes from the repricing of risk is borne out by the rise in the default spread for bonds, with the Baa default spread widening by 0.17%, and the increase in the perceived riskiness (volatility) of stocks, with the VIX posting its largest weekly jump ever, in percentage terms.
The Repricing of Risky Assets
When the price of risk changes, all risky assets will be repriced, but not by the same magnitude. Within mature markets, you should expect to see a bigger drop in stock prices at more risky companies than at safer ones, though how you define risk can affect your conclusions. If you define risk as exposure to the the precipitating factor in the crisis, I would expect the stock prices of companies that are more dependent on China for their revenues to drop by more than the rest of the market. Since I don't have data on how much revenue individual companies get from China, I will use commodity companies, which have been aided the most by the Chinese growth machine over the last decade