Bill Rempel, a.k.a. NO DooDahs!

Why the last Friday in April of 2005? Urrrp! Why not? It’s an arbitrary and capricious decision, but it’s my blog. I wrote about the “undeniable short setup” a couple of months ago, and backtested it on the SPY. It was an apparent sign of short-term weakness but certainly wasn’t a reliable signal to go short the SPY. Given that the index shows more mean-reversion than individual stocks, I thought it might be worth looking at some history. If this post gets good feedback, it’ll be part of a series.

If we were to apply those rules, stated as “a week with (1) light volume, (2) a six-week high, and (3) a close below the open,” to stocks (with an arbitrary liquidity minimum applied) as of the last Friday in April of 2005, we would find 65 stocks met the criteria. In order to narrow down the charts a little bit, I’m going to (again, arbitrarily) limit the list somehow to ten stocks. To do this, I am going to pick the ten with the highest six-week price increase (including the signal week’s returns). I considered picking the ten with the highest 52-week returns, or some other measure, based on the interpretation that this setup is designed to short short-term weakness in longer-term strength, but eventually decided to confine the scan to the six-week period. The cutoff turned out to be +7.3% over six weeks, including the down week, which most would probably consider to be a large gain in a short time.

The stocks returned by this scan are CAI, CELG, CORI, EIX, GPN, LPL, PDC, TALX, VTS, and XPRT. I’ll show every chart here, without annotation. You can spot the “signal” week by looking at 2005 and the last candle in the month of April, it will be filled with red on the chart. The proposed entry is probably at the open on the following candle, but I suppose a quick trader could scan intraday/week and enter the short at the close of the signal week.

Shorting CAI got you three weeks of pain, but if astute, you could have covered in the money … in a month or so.

The CELG chart speaks for itself. A short would have been profitable with a tight stop in the first couple of weeks, but not after that.

Any short of CORI on that signal was out of the money for most of the following six months.

EIX gave you some minor opportunities for gains before telling you “you’re dead, shorts.”

GPN could have been covered with a profit in the next two weeks, it actually repeated the “signal” in the following week, but it’s hard to say that the signal was a sign of anything.

If you shorted LPL on that signal, it broke your arms and blew out your stops.

XPRT is one that worked! Hooray! A profitable short on this signal!

PDC is another one that worked! Well, depending on your stop. From $13 to $14 before going down is a scary short ride.

TALX shot you in the back of the head immediately after you shorted it based on the signal.

VTS gave you one week of pain, a week where you could have covered profitably, and never looked back.

So, out of ten “undeniable short setups,” only one really looked undeniable. It looks like this might be a promising technical screen for long plays, however. A week of weakness in a pattern of longer-term strength … hmm …

In terms of Stockcharts’ scanner, it looks like this:

[type = stock] and
[[exchange is nyse] or [exchange is nasdaq] or [exchange is amex]] and
[sma(30,volume) * sma(30,close) >= 1,000,000] and
[max(5,high) = max(30,high)] and
[close < 4 days ago open] and
[sma(5,volume) < sma(30,volume)] and
[roc(30) >= 7.3]