The indicator looks at the component stocks of the S&P 500 and computes how many points gained/lost were lost, and also how much of the volume flowing into up/down issues went into down issues.

Over the past 10 days, the average of those two figures is 73%, meaning about 73% of the points gained or lost in the component stocks were lost, and about 73% of the volume flowed into issues down on the day.

There have been four other days that match or exceed this reading – 7/22/02, 10/9/08, 7/2/10 and 8/4/11.  They were each within days of vicious market bounces.

 The specific Fed Funds target doesn’t have as much history as the Discount Rate, and the two move in tandem almost without exception; when one is rising or falling, so is the other one.

Using the Discount Rate back to 1934, we can see how the S&P 500 fared during the worst six months of the year, from the end of April through the end of October, under various interest rate scenarios.

Common perception is that rising rates are bad for the market since borrowing costs are rising and forces are aligning to slow growth and inflation.  Combining rising rates with bad seasonality should be a deadly one-two punch.

The table below shows the S&P’s performance over the six months starting in May when the Discount Rate was rising, falling or holding steady over theprevious six months.  So if the Discount Rate was higher on the last day in April than it was on the first day of the previous November, we would consider rates to be rising.

We can see from the table that the worst scenario was, indeed, rising rates.  When the Discount Rate was rising, the summer months were positive 54% of the time over 24 years with a median return of only +0.6%.

Even more notable, the S&P’s maximum loss at its worst point during the summer was greater than its maximum upside.  Only 11 of the 24 years managed to gain more than they lost during the summer.

If the Discount Rate was falling and monetary policy was loosening, then the S&P performed better.  It was positive from May – October in 65% of the 20 years, had median return of +2.0% and its maximum gain was higher than its maximum loss…marginally.  In 12 of the 20 years, the S&P managed to rise more than it fell during the summer.

Part of the problem in some of these years is that the Fed tends to loosen as the economy is weakening, and often that has coincided with some very poor stock markets.

The last scenario is when the Discount Rate was steady over the previous six months.  Perhaps because it coincides with periods of relative calm, this is when the market did its best.  The S&P was positive during the “worst six months” nearly 70% of the 28 years and averaged a respectable +3.7% return.  Its maximum gain during those spans was well above its maximum loss.  It managed to rise more than it fell in 17 out of the 28 years.

If the Discount Rate was steady and below the long-term average of 4%, then May – October was positive 75% of the time, with a median return of +4.9%.

If the Rate was steady but above 4%, then the summer was positive only 57% of the time with an average return of +2.1%.

So with the Discount Rate steady and holding well below average, it looks like we have the best possible scenario.  At least in terms of it being the worst six months of the year.

What about the best six months, from November through April?

Rising rates

Avg Return:  +6.6%, 81% positive

Max Loss:  -6.6%

Max Gain:  +9.1%

 Falling rates

Avg Return:  +5.6%, 86% positive

Max Loss:  -2.0%

Max Gain:  +11.2%

Steady rates

Avg Return:  +5.0%, 65% positive

Max Loss:  -2.7%

Max Gain:  +9.2%

By far, the best scenario was when the Discount Rate had been falling over the summer.  In those cases, the S&P was positive during the next six months 86% of the time and with a maximum gain that averaged five times greater than the maximum loss.

The worst performer, at least in terms of average return and consistency, was when rates had been steady.

Bulls could hope that the Fed lowers rates before November, but it’s hard to imagine them going any lower than they already are.

We’ve looked at volume into and out of those funds in the past, and it has been mostly effective…except lately.  So what about the actual assets in those funds?

We took a look at the top bearish ETFs that had more than 500,000 shares in average daily volume over the past 3 months.  There were 19 of those.  Then we chose the ones that have released at least two years’ worth of daily asset levels.  That left us with 13 funds.

Over the past few years, the assets have formed a fairly defined range, between $6 billion on the lower end and $11 billion on the upper end.

The assets are clearly getting low, currently totaling $6.7 billion.  The history at the lower end is too short, however, to really determine if this is an adequate contrary indicator on an intermediate- to long-term time frame.  The funds just haven’t been in existence long enough to find out.

The chart below shows the Bull Ratio for all the index funds.  This is the total assets in the long funds for the S&P 500, Nasdaq 100, DJIA and Russell 2000 divided by the assets in the inverse funds for those same indexes.

As of Monday, there was $4 invested in the long funds for every $1 invested in the inverse funds.  The only other time since 2004 that matched this level was May 2, 2011.

As of Wednesday, total assets invested in the inverse funds at the Rydex family fell to $291 million.  That ties it for the lowest all-time, along with January 30, 2001 and May 2, 2011.  Both dates were good times to own those funds.

The S&P 500 went the entire month of January without a meaningful drop, so traders are not seeing much point in hedging against seeing one anytime soon.

The chart below shows the volume in inverse ETFs.  We can see that the other times this hedging volume was so low, stocks got hit soon afterward.  Even if we compare this volume against total composite NYSE volume, it’s almost exactly the same picture.