Pandering to the peanut gallery, let’s take a look at a few bearish arguments for equity markets right now. First up, a simple rate of change indicator from On the Money blog called the “Chart of the Century”. This is simply the 10 day Rate of Change of the Dow Jones using a monthly interval.
ROC and Roll
The charts I’m sharing with you below are slightly different, on a weekly basis, using a 40 week RoC, but they show the same basic picture:
The charts are beautifully simply and speak for themselves. This basic indicator has been an early warning signal for almost all major market declines, as well as plateaus - including the 1987 crash. Obviously the spike up in late 2009 is a result of the deep decline in 2008. It reminds me of the more long term chart of 10 year rolling returns I shared with reader in March 2009: Revisiting The Long Term Bullish Case For Stocks. That simple momentum indicator was a major signal that we were at a major inflection point. And that is what the “Chart of the Century” is suggesting as well.
The 2008 credit crisis was truly unprecedented in many degrees. And today we are feeling aftershocks as financial stress indexes rise once more - thanks to the ongoing Euro crisis.
While the TED spread has reversed in recent days, there are other measures that show the credit stresses are building up. Of these, the spread between Moody’s Baa Corporate rate and the long term US T-bill has a strong negative correlation to the stock market. According to John Kosar, Director at Asbury Research, since 2005, the spread has been 70% inversely correlated to the S&P 500 index and 92% inversely correlated since 2009.
I decided to take a much longer term view of this spread using the constant maturity 20 year T-Bill rate. Moody’s tries to keep the term at least 20 years so I thought the 20 year constant maturity would be a good fit. Others use the 10 year but I really don’t think it makes much of a difference.
Instead of comparing the spread to its 50 day moving average, I plotted a 100 day RoC to better show the change over this time period:
The chart looks like a financial seismograph! But even more interesting, are the predictive qualities of the indicator. I don’t think that is a coincidence. Since the majority of stock market returns come as a result of “animal spirits” (P/E contraction/expansion or sentiment) any difficulty that firms have in finding cheap financing will eventually filter down from the bond market to equities.
Most of the time this is an early warning signal, months ahead of time. For example, it passed the important 0.50 level in March 1998, months before the S&P 500 fell 18% in the fall of the same year. And in 2007 it once again breached 0.50 in April
On June 11th, 2010, the 100 day RoC of the Moody’s BAA and 20 Year T-bill spread reached 0.56. Since then this indicator has dropped off slightly but it is still relatively elevated at 0.48 (the latest reading for June 28th 2010). And perhaps more importantly, the spread itself continues to widen and sits currently at 2.03%.
Duelling Quant Models
You might recall that back in March, QuantDNA issued and reiterated an exhaustion pattern signal. According to them, this model had been successful in guarding investors from past declines by issuing similar sell signals in October 2007, April 2000, July 1999, July 1998, October 1997, January 1992, August 1991, June 1990 and finally, the important one, August 1987.
The signal was a bit early since the S&P 500 topped out on April 26 at a high of 1220. But depending on how far the decline is, this was either a good or excellent signal.
Allowing us to finish off on a slightly upbeat note, François Soto, the founder of Emphase has shared his Credit Complacency Crash model. According to him, he has found a method to not only predict every single major crash since 1965 but also each trough.
I did notice that in previous charts (updates for April and May) the model was unable to flag the 1998 ‘crash’ but in the most recent update (for June - graph not shown) the model does provide a signal to protect against the 1998 decline. I asked Francois about this and he replied that the tdata series he used was updated due to FASB adjustment for all years and the parameters had to be adjusted. He added, “With some slight adjustments it captured the 1998 crash that way. But there is no signal for 2010 as credit conditions and neither debt did not pick up since the 2007 recession yet… Hope this helps clarify.”
The important take-away is that according to this model, there is not upcoming crash. Don’t you feel better now?
Enjoyed this? Don't miss the next one, grab the feed or