Insights & Advice


1987 Remembered

August 29, 2007

Back in June I wrote an article I entitled “Return Compression is the Enemy” where I outlined that asset classes (large cap, small cap, value, growth, etc.) and sectors (technology, financials, etc.) all had been performing very similarly and as a logical result it had become difficult to outpace the stock market indices (if everything is performing the same, how do you outperform?).

In an effort to use this data as a forecasting tool I looked at past periods where this type of return compression was present to get a glimpse of how the market had acted afterwards.  I wrote that “there has been a lot of return compression over the last 12-months. Over the last year the percentage of S&P 500 stocks that outperform that index has dropped from 60% (about the highest since 2002) to 41.2% (the lowest since about 2000).  So while that was clearly bad news for investors trying to beat the overall market last year, our job now is to determine how to use this current knowledge.  In other words, what does this type of return compression mean for the outlook of the stock market?  Unfortunately the signals are mixed. 

Historically (and remember, we are only looking at one indicator out of dozens and dozens) when these readings dip below 43% the market tends to follow with good strength.  But there are important exceptions.  The most recent was the crash of 2000 (when this compression ratio last dipped below 43%) as well as the just before the crash of 1987.”

In that article I had argued against the likelihood of a 20%-type crash, instead continuing to look for a typical once-per-year 10% drop, probably sometime in the summer.  So far that is what we have seen.  But, since it is very likely that the markets will re-test their August 16th lows, I do have to take into consideration that about one-third of the time that there is a 10% drop it turns into a 20% drop. 

Typically a 20% drop isn’t the end of the world (although sometimes it feels like it); especially because that type of drop, on average, takes almost a year (332 days), so we have plenty of time to adjust our portfolios.  However, in a 1987-like drop (and from July 19th to August 16th we went through a mini-1987) you just can’t systematically get out of the way – damage will be done to your portfolio.

Now the good news is that “circuit breakers” have been put into place so that the market will not sustain a similar 2-day 20% meltdown.  However, with the recent removal of the uptick rule (up until very recently you could only short a stock after it ticked up – known as the uptick rule; now you can do it whenever you want and that helps to spark and snowball violent downward movements) as well as all the quantitative hedge funds all doing the same things when the market goes down (their models have them selling automatically as the market goes down, exascerbating the problem and actually hurting stock prices and themselves – it’s part of the quant problem nobody seemed to understand had to be part of the model), there certainly is the potential for steep downward spikes in a short period of time (think 1997-1998 Asian Contagion – a 7.2% drop on October 27th, 1997 in US stock prices and a 20% drop over the fall of 1998).

Ned Davis Research put together an interesting table comparing then and now.  There are twenty similarities and eleven differences, as follows:


















The Tape

















   S&P 500 first drop from high










19 trading days



18 trading days










   Recovery rally


5.2% (to 8/24 high)














   10% correction drought

second longest



third longest

   (at peak)

















   Divergences at peak

Percentage of new highs,







90-day advance/decline ratio,







Dow Utilities
















Underperformed into peak







and after

































   US Dollar Index


Weakened into market peak,







maintaining long-term downtrend













   Crude Oil


January-July rally, then weaker













   CRB & Gold


Maintaining Uptrends














   CPI less core CPI


0.1% for July



Same for peak month, August 1987










   Consumer Confidence

112 for July



111 for August










   US budget deficit


Narrowed to $166B for July


Narrowed to $169B for August










   Baa bond yield momentum

Reached bearish mode six weeks

Reached bearish mode 11 weeks




before peak



before peak










   Composite global


Yield momentum positive at peak



   interest rates

















   Advisory service sentiment

72% bulls when market peaked


74% bulls when market peaked










   Net retirement of equities

Positive for 12 straight quarters


Positive for 14 straight quarters




through Q3



through Q2




























   Economic Expansion

Sixth year since recession


Fifth year since recession










   Presidential term


Seventh year of two-term Republican












   Fed Chairman


Second year of a new term


First year of a new term










   Trade tension


With China



With Japan










   Strategy influence


Quant trading



Program trading




































   Y/Y CPI



2.4% for July



4.3% for August










   Discount rate


Cut after peak, first following


Raised after peak, first following




tightening cycle



easing cycle










   Y/Y T-Bill yield


Falling yields starting in June


Rising yields starting in July










   Profit margins


7.6% in Q1



3.8% in Q2










   Y/Y S&P earnings


up 16% in Q2



down 2% in Q2










   S&P P/E


16.9 in July



22.9 in August










   T-Bill yield /earnings yield

0.9 at peak



1.4 at peak










   10-yr. yield minus 3-mo. Yld

0.2 at peak (flat yield curve)


2.7 at peak (steep yield curve)










   Real M2 minus Ind. Prod.

up 5.4% in July



down 8.3% in August










   Uptick rule

















   Circuit breakers








The last item indicated in that chart – circuit breakers – should protect us from a 20% drop in the market in any given day.  But it doesn’t rule out a 5%+ drop or a crash over a very short period of time.  Still, should the stock market continue to go south, I don’t see that as a high probability (well, I guess “high” is relative – historically speaking there is a one-in-three chance that the current downtrend will hit a “crash” level of 20%).  The higher probability event (we’ll go ahead and give it a nearly two-in-three chance) is that the stock market will come close to re-testing the August 16th market lows (1,370 points on the S&P 500) sometime over the next two months and then prices will rebound.

So what has to happen to get the crash?  Well, certainly the elimination of the uptick rule and the advent of quant trading is not going to help.  But the bigger problem is going to be the economy and whether or not we get into a recession.  So far there is a measured recession in the housing market, but it has not (yet) affected to the broader economy to the same extent.  As long as economic recession is avoided then a continuation of the market’s correction will most likely simply retest recent lows and end up being a standard once-per-year (technically about once every fourteen months) 10% correction that everyone forgets about months later.