Market recoveries - historical perspective

 I will be the first to admit that the strength and staying power of the current stock market rally, both in the US specifically and in global markets more generally has caught most people, including myself, by surprise.

 While a bounce post-Iraq was to be expected (when the markets came within a whisker of the October 2002 lows), the extent of the rally( S&P 500 is up more than 25 per cent from its March 11, 2003 low) and more importantly its staying power has been unexpected.

 While I still expect the markets to correct as we enter the seasonally difficult period of late September- October and feel that the market has gone up too far, too soon, it makes sense to look at this rally through the lens of history.

 What can history teach us about past market recoveries from severe bear markets? Has this rally been more powerful and quicker than the historical norm? How long can it be expected to continue if it follows past patterns?

 In this light I came across an interesting study done by the Leuthold group examining exactly these issues. It examined 100 years of evidence of stock market rallies post a bear market low and came to some interesting conclusions. Unfortunately the study is US centric but still throws up some interesting results which should have broad relevance for markets globally.

 The study traces 22 bear market lows over the past 100 years in the US, with a bear market being defined as a peak to trough decline in the market averages of 20 per cent or more.

 The study points out that the average peak to trough bear market decline over the last 100 years is 37 per cent, with the median decline being 34 per cent. Eight of the 22 bear markets declined by more than 45 per cent.

 The recent bear market of 2000- 2002 saw the third biggest decline of the S&P 500 in history (-49 per cent), (assuming of course we have seen the lows for this cycle). The biggest market decline was during the depression of 1929-32, when the market declined by 86 per cent.

 The average bear market decline lasted for about 19 months, with a median duration of 20 months. The markets tended to be in a bear market mode for approximately 34 per cent of the time over the past 100 years.

 Looking at recovery statistics the study comes to the following conclusions:

 

 

The majority of stock market gains in a recovery occur in the first year. In fact the average stock market gain in the 12 months after hitting a bear market low was 47 per cent.

Typically, additional gains can be expected in the second year of the recovery, though these gains will not be as significant as the first year. There were only three instances out of 22 when the market actually fell in the second year of a market recovery post a bear market low.

The third year of the recovery is typically not positive, more than half the time the market declined in the third year of the recovery.

 Also if you look at the monthly dispersion of returns, not surprisingly the maximum returns are achieved in the first month of the recovery, when the markets typically rise by 13 per cent, but there are consistent gains in cumulative performance well beyond the first few months.

 In fact, performance does not appear to level off until month 21. Thus even if one were to miss out on the first month or two of the recovery, significant returns can still be had.

 Looking at all of the above data, the obvious next question is how is the current rally tracking in comparison with the historical norms of the past 100 years. Is the popular perception that this rally has gone too far, too soon justified on the basis of past history or simply rooted in emotion and fear.

 If you accept October 9, 2002 as the bear market low for this cycle(S&P 500 closed at 776.76, and take the March 11low of 800.73 as a test of the October level), then one can track this recovery in comparison with the average of the past 100 years.

 The study does exactly this and comes to the following conclusions:

 As of August 19, 2003 the S&P 500 had gained 29 per cent from the October 9 bear market low in a period of ten and a half months. Surprising to most would be the observation that this recovery is actually tracking well below the average of past bear market recoveries.

 Going by the evidence of the past 100 years, markets should be up an average of 40-44 per cent by the eleventh month from the bear market low. Thus the 29 per cent rise we have seen so far is well below the norm, and does not support the common perception of the market having risen too far, too soon.

 This notion is reinforced if you consider the fact that recoveries are normally much more powerful when you are coming off severe bear markets like the one we have just experienced.(First year gains following big bear market declines of more than 45 per cent, average 68per cent).

 The message from history is clear; this market has clearly more to run and is far from over extended either in terms of absolute performance or duration.

 If we extend this analysis out longer to look at a more complete recovery cycle, than based on history, the S&P 500 should peak out around 1,250 near the second year of the recovery.

 This implies another 25 per cent upside from here and continued positive performance till August