Institutional Distribution, Part 2

1) Distribution clusters appeared near every major market top.

2) The reliability of these distribution clusters in signaling a top was
questionable.

The most common way that intermediate-term traders use distribution days is in
helping them identify market tops. Today I will actually quantify their
reliability so that you may better understand the significance of a distribution
cluster occurring in an uptrend.

To help me quantify this I created a strategy that met specific criteria. The
strategy was tested over the period of the last 15 years.

I would look to sell short the S&P 500 under the following circumstances:

1) The S&P 500 closes above it 200-day moving average. (Remember, we are looking
for tops here.)

2) Sometime within the last 12 days the S&P 500 closed within 1% of its 200 day
high. (Again, confirming we are near a top.)

3) A distribution cluster occurs. (Same definition I used in ShowMe Study on
July 20 — 4 distribution days within last 12 trading days.)

If the above conditions exist, the strategy will look to sell short the index
(anticipating a top) and then cover the trade 20 trading days later.

Here’s a few statistics based on the above strategy:

Total # of Trades — 58

Winning Trades — 21

Pct Profitable — 36%

Profit Factor (total gains divided by total losses) — 0.47

For those of you who are not accustomed to analyzing strategy performance, let
me do it for you. These results are abysmal. Terrible. Awful. (I also looked at
a 40-day holding period and they were even worse.) In fact, the results are so
bad, we could reverse the strategy (buy under the same conditions and sell 20
days later) and get the following results:

Trades — 58

Winning Trades — 37

Pct Winners — 64%

Profit Factor — 2.13

Now those are some fairly impressive results. Personally, I would never trade a
strategy that has you sit in a trade for 20 days with no trade management
parameters laid out, but the point should be obvious:

Distribution clusters on their own are a very poor indicator of a market top.
While they DO occur at nearly all market tops, nearly 2/3 of the time over the
last 15 years the market has merely pulled back before advancing further. Under
most circumstances, you are better off looking to buy these pullbacks rather
than joining the panic and selling.

Price and volume action in the major indices may be a useful tool in determining
the health of the market, but I would certainly be careful about overemphasis
when it comes to analyzing potential market tops. The clues that price & volume
are able to provide should always be viewed within the context of what other
market indicators are saying.

Best of luck with your trading,

Rob

RobHanna@comcast.net

P.S. If anyone with Tradestation 8.0 Build 1869 or later would like a copy of
the July 20 ShowMe Studies or today’s Strategy, just drop me an email.