Short-term Outlook (1-5 Days):  25% Bearish  From Apr 13, 1194 SPX

Equity Indicators - Updates and Extremes

Small Options Traders

We've looked at extremes in put/call ratios several times lately, and each time I hear back that we have to discount these readings because:

*  it was options expiration week

*  it was all due to one or two stocks

*  it was mostly due to dividend-capture strategies

*  it was mostly due to stock-replacement strategies

*  traders weren't actually buying a lot of calls, it was mostly selling-to-open volume

*  traders weren't buying calls, they were just not buying puts

I'm not going to refute each point, as I've touched on a couple of them over the past few days and some are just without merit.  Instead, I want to look at the ROBO Put/Call Ratio, which is about as pure an option indicator as we're going to find.  It will address most of the points above.

The indicator only looks at the smallest of options traders, those trading 10 contracts and less.  And it only considers orders for buying-to-open.  So we have a really good handle on who is doing what.

I feel exceptionally comfortable with the assertion that the vast (vast!) majority of those trading less than 10 contracts is not involved in complex portfolio maneuvers; they buy a call option to speculate on a rally, and they buy a put option to speculate/hedge on the downside.

So let's look at this week's ratio:

Based on other options data, it's pretty obvious that Friday's reversal put a damper on the speculative mood, but even so the rush into call options (and continued lack of interest in put protection) is some of the most egregious we've seen since 2000.

Small traders spent 43% of their total option volume on buying call options, the most since October 2007.  Less than 15% of their volume went to buying protective put options.

The table below outlines each week (excluding the year 2000) that the ROBO Ratio was about equal to the current one, along with the S&P 500's performance going forward.

There was really only one time, in November 2003, that the surge in speculation didn't precede a correction (in terms of any further upside being erased away in the days or weeks ahead).

On Friday, we looked at the Intermediate-term Indicator Score, and noted that every time we've seen a similar spike, high-beta indexes like the Nasdaq 100 were about to under-perform the broader market.

The same is true after the extremes in the ROBO Ratio mentioned above - after every occurrence, the Nasdaq 100 under-performed the S&P 500 as the "risk trades" lost steam and speculation dissipated.

Large Decline In A Momentum Market Near A High

Friday marked the largest decline in two months for the S&P 500.  It's no secret that the index was also trading at a fresh 52-week high just the day before, and has enjoyed remarkable momentum over the past two months.

This kind of thing gets traders all excited about the potential for a trend change.  Curiously, though, that rarely happens.

The table below shows all occurrences since 1928 when the S&P 500 went longer than a month without closing below its 10-day moving average, closed at a new 52-week high, then suffered its worst one-day loss in at least two months.

The last two columns in the table highlight how long it took (and what kind of pain investors suffered) from the day of the large one-day loss until the S&P recovered to close at a new 52-week high again.

The average returns and percentage of time the S&P was positive weren't that great, but they weren't too terrible, either.

There were really only 3 times out of the 12 that marked a major peak, taking longer than a month and more than -3% loss before recovering.  The median number of days to hit a new high was only 11 days, with 8 of the precedents making it back in under a month.  The median drawdown (i.e. maximum loss) was smaller than -2%.

I've looked at this phenomenon other ways, too, such as when sentiment is overly optimistic (e.g. the Investor's Intelligence Bull Ratio above 70%).  The results were pretty much the same - out of the five instances, it took a median of 12 days to close at a 52-week high again, and the median drawdown to get there was -1.9%.'

Only once did the large one-day loss mark a major peak, while twice it led to a market that went pretty much nowhere for months on end, and twice it led to only a very minor short-term dip before another rip higher.  No conclusions to draw there.  The dates were 2/5/76, 10/11/83, 1/8/86, 1/23/87 and 1/27/04.