By Jonathan Bernstein
Yes, it's true, "headline risk" explains volatility, at least on the surface. But no other news driven market of recent decades had anything like the volatility we've sometimes seen since 2008. Think 1989-91, which included Drexel's bankruptcy, and Iraq's invasion of Kuwait. Or even 1987, where volatility reached extreme levels, but didn't stay that way for long. Or 2000-01, where the Nasdaq had some very volatile days, but the other indexes weren't affected nearly so much.
Only the 1930s saw so many big daily fluctuations, because investor emotion plunged and soared in the extreme from day to day, just like it has recently - only much more so then than now.
Volatility Was Greatest in the 1930s, Only 2008 Looked Anything Like It
If you don't believe me, here are some numbers. In the whole period of 1989-91, the Standard & Poor's 500 (SPX) had a total of only five - five! - days in which the daily price range exceeded plus or minus 3%. In 1974, the S&P had only nine days in which the S&P moved more than 3%. Pretty gentle, considering that the market fell about 45% in 1973-74. In 1987, the total of high range days was also nine. However, when we look at 2008, there were 42 such days, just for one year. Quite a difference.
Since the SPX began in 1957, I use the Dow for earlier history. Of the 20 biggest daily percentage losses in the Dow's history, nine of them occurred between 1929 and 1933. And of the 20 biggest daily gains, 15 of them occurred between 1929 and 1933 - and one more in 1939.
Of course part of the reason for the big daily percentage gains (and losses) of the 1930s was that large point fluctuations had an exaggerated effect on the calculation of daily percentage moves, since those fluctuations were measured against a smaller post-Crash base. On the other hand, the '29 Crash took the Dow down almost to turn of the 20th century levels, and we didn't see many big daily percent fluctuations in the years 1900-1928. Even though that period included the Panic of 1907 and the post WWI market break of the early Twenties.
And how about this? To my eye, it looks as though nearly all of the five-year period, 1930-35, the 200-day moving average of daily ranges held firmly above 1.5%, and even touched 2.51%. These days 3% is considered a big daily move but then that was not that much more than the average for at least one 200-day period.
Investor Emotion Is Exaggerated When Financial Collapse Is Feared
So what links the 1930s markets and more recent markets to today's volatility extremes? My answer is that in the 1930s, investors feared that things might all come undone, and that kind of fear has surfaced occasionally since 2007. As we know all too well, when people are afraid of a U.S. credit crisis (2008) or European contagion, they just sell all risky assets. And when investors think the danger has passed, buying panics happen instead. Investors really hadn't acted like this since the 1930s.
I read the news from 1930 blog every so often, because it illustrates at least some similarities of market psychology then and now. Each of the blogger's entries centers on excerpts from Wall Street Journal articles written on one given day in the 1930s to show readers what drove markets at the time. And a recent entry from that blog, which includes WSJ snippets from September 25, 1931, could almost be ripped from today's headlines. Fears about a break in sterling, which was an issue not so long ago. Fears of the U.S. losing its foreign exchange reserves, which then consisted of gold held at the New York Fed. Big fluctuations in silver, driven in part from demand swings in India. No wonder the Dow fell 7.1% that day.
If you scroll down you'll see the entry for September 24, 1931, when stocks soared 6%. That's right, the Dow notched a 6% up day on the 24th, followed by a 7.1% loss on the following day. We didn't see numbers like that in 2011, nor did we so much even in 2008. History hasn't repeated itself, but it has rhymed.
What news drove the September 24th buying? Again, the headlines look eerily familiar - other than the re-opening of the London Stock Exchange after a two day shutdown, at least. The American Stock Exchange (remember them?) lifted its ban on short selling, foreign currencies rallied, U.S. workers took wage cuts. The more things change.
Just as an aside, bank analyst Thom Brown recently said on Bloomberg that the banking sector's beta is now about two; ie., if the S&P moves 1% up or down, bank stocks will move twice as much. My explanation of this: When collapse is feared, bank collapse is feared the most. And when panic lifts, investors pile back into bank stocks. It's hard to believe that not too many years ago bank stocks were considered staid dividend payers.
Deposit Insurance Keeps Volatility Down, At Least When Compared to 1929-33
Of course the bank bailouts and massive liquidity injections (of which there is much to dislike) helped the markets regain their calm within months of the Lehman bankruptcy. Just as important, though, is the fact of deposit insurance. These days, a frightened investor can go to cash and park it in an FDIC insured account, something we now take for granted. But until the FDIC began offering insurance in 1934, investors had no such comfort. Hundreds of banks blew up each year in the Twenties and the flow accelerated into the Thirties. And back then when banks failed, depositors often came out with nothing. One could find safety only by stuffing cash or gold in the mattress.
If we go back to our numbers on daily volatility we can also see that big percentage moves got noticeably rarer starting in 1934. Coincidence? I don't think so.
Due to the FDIC, and Treasury's new money fund guarantee, the 2007 money market panic that broke out after the Reserve Fund "broke the buck" was a walk in the park compared to Depression-era liquidity panics. A bad day in the 1930s could vaporize your savings, even if you had them in the bank. On the good days, fear of ruin would give way to fear of missing the mother of all bottoms, since investors well knew how cheap stocks were then. Today we see at least an echo of this.
What Traders Should Do
While volatility has fallen off in recent weeks, I don't think it's gone for good. No matter which way the markets go from here, volatility should rebound; it's hard to imagine long-term calm returning to the markets until we resolve the overhangs of public and private debt, here and in Europe. Therefore, those who like to play volatility should think about buying volatility while the VIX sits in the teens or low 20s. If the trade works, lock in partial gains when you can and hope to sell the rest of the position near the VIX resistance levels of recent months. Of course if this rally just keeps going (and some recent rallies lasted longer than I thought they would) the trade breaks even at best.
All the related instruments I know of have their drawbacks: The VXX has problems common to leveraged ETFs ... VIX futures entail contango risk at rollover time, and theta eats away at option positions. But I do like the risk/reward here for volatility longs. Conversely, shorting the VIX near last fall's highs worked out well, and could do so again.
Disclosure: I have no positions in any stocks mentioned, but may initiate a long position in VXX over the next 72 hours.