How Long Can This Market Creep Up? Lessons From The Past

Includes: SPY
by: Patrick Chu

As of today, February 28, 2012, the S&P 500 index has gone 41 trading days in a row without a decline of 1% or more close-to-close, a condition I am calling a "calm" rally (easier to type than "low volatility rally"). This calm rally started on December 28, 2011, has returned 9.8% so far, and the largest decline was -0.7% on February 10.

I was wondering how often calm rallies of this length had taken place in the past, so, armed with a spreadsheet and historical S&P 500 quotes back to 1980, I generated the chart below for all "calm" rallies (again, defined as a rally without a decline of greater than 1%) of 41 trading days or longer.

Calm rallies of 41 days or more since 1980Click to enlarge
(Click to enlarge)

(Note that the latest calm rally has not yet ended).

Distribution of Calm Rallies by Decade

From 1980 through 1985 we saw no calm rallies. We saw two calm rallies in quick succession in 1985, then two more in late 1988 through mid 1989.

Late 1991 through the end of 1996 was the heyday for calm rallies, including most of 1995, with a six-month calm rally returning over 18% and another one right after returning almost 12%.

After a nice 3 month calm rally in 1996 that returned over 15%, we didn't see another calm period until late 2003, almost 7 years later. Obviously the Internet bubble period was anything but calm, with sharp moves both up and down even before the eventual crash. After 1996, the drought column tells us that almost 1800 trading days passed between 1996 and the next calm rally in 2003.

Starting in 2003, we saw short calm rallies late in the year for every year from 2003 to 2006, with the longest calm rally taking place the latter half of 2006 for a gain of almost 13%.

After 2006, we did not see another calm rally until the current one starting in the very last days of 2011. At 1300 days, the current rally came after the second longest "calm rally drought" we've had since 1980.

How Subsequent Returns Are Calculated

(You can safely skip this section if you don't care)

The month, quarter, and 6 month return periods are defined as 20, 60, and 120 trading days rather than dates on the calendar. Returns are based on the S&P 500 index, not the SPY ETF.

The "subsequent returns" are calculated starting with the closing price of the day the drought ends, that is, starting with the closing price of the day of the greater-than-1% decline. Example: if you bought the S&P 500 at the close of the day the calm rally ends, and held it for 20 days then sold it at the close of that day, you would get the "monthly" return.

Note: The monthly returns in the table are probably overstated, because if you were holding the S&P 500 while the rally was in progress, your monthly return would be less because the monthly numbers in the table EXCLUDE the 1%+ decline at the end of the rally.

It appears that in every case after the rally ends, the monthly and quarterly returns afterwards do not match what you made during the rally. On the other hand, the returns do seem to be generally positive 60 days after the rally ends, at least in every case since 2000. Of course, markets at each moment in time will differ, and our returns will depend on what is happening with the macro picture, i.e., events in Europe and whether quantitative easing is expanding or ending.

After the Drought

Our current calm rally comes after a drought of over 5 years, so the calm rallies starting January 1985 and October 2003 can serve as "key races" (as they say in horse handicapping circles), since those calm rallies also came after droughts of over 5 years.

Those drought-ending calm rallies returned 11% to 13%. The 2003 calm rally lasted 65 days, and the 1985 calm rally lasted a whopping 112 days. Both rallies came after the market recovered from very volatile periods in the stock market. In each case after the drought was broken and that calm rally ended, another calm rally started less than a year later.

Predictions and Analysis for the Current Calm Rally

It appears that, except for periods of very low volatility in the mid 1990's, most calm rallies can easily generate double digit returns. While the above table is somewhat misleading because I only show calm rallies of 41 days or more (making it look like every calm rally lasts a minimum of 41 days, which is not true), a calm rally lasting 60 days is quite possible, especially after breaking a drought of calm rallies. Absent earth-shattering news, we could easily be in for another month of low volatility and creeping upward prices.

In addition, once a multi-year drought is broken, more calm rallies tend to follow. It's possible that we could see another period later this year of low volatility, despite the rising price curve in later-month VIX futures.

The cliche is that markets climb a "wall of worry". I've rarely seen this more the case than in this calm rally. Many posts here on SeekingAlpha and on, the bears, who far outnumber the bulls, are expecting a crash any day now. The talking heads on CNBC tell us that many funds and retail investors have missed this train.

If that's so, then where's the money coming from? My guess is the ECB's LTRO program. A few hundred billion dollars is enough to keep the boat afloat in a thin volume market. And guess what? As you know, tomorrow starts the second round of LTRO, which just might continue this calm rally for another month or longer.

In any case, calm rallies of the recent past do not support a "crash" scenario. Since 2000, the biggest rally ending day was a decline of -1.87%, on January 19, 2006. The other declines since 2000 were -1.36%, -1.36% and -1.12%. After recent calm rallies have ended, the subsequent monthly and quarterly returns were all slightly positive, and with quarterly returns of 3-5%, not too shabby either, on top of the double digit gains from the calm rally itself.

Of course, a chart like the one above has limited predictive powers where exogenous factors like Europe are in play. If a European bank goes under, for example, then anything can happen. Like the effect of The Mule in Issac Asimov's Foundation Trilogy, when you have unpredictable external factors influencing the outcome of your model, then all bets are off.

Disclosure: I am long VXX. I own long-dated AAPL puts at different strikes. No position in SPY.