Day Trading With Stochastics: by John F. Kepka
Day Trading With Stochastics: by John F. Kepka
Day Trading With Stochastics: by John F. Kepka
Kepka
I have been using the Stochastics oscillator "K" and "%D" for well over a year (RE: May and Sep 1984
issues of Technical Analysis of Stocks & Commodities magazine). I use Stochastics, (in conjunction with
other indicators), with futures intraday data from half-hourly down to 1 minute bar charts. Initially, I tried
to use the Stochastics indicator as outlined in Technical Analysis of Stocks & Commodities, but I quickly
learned that the "clearly illustrated" patterns were largely non-existent in real time intraday trading. By
following the general rules, the "Set-up" rapidly became expensive. Even "shooting from the hip" was
better.
After considerable experimentation, I concluded that I definitely preferred the regular "K" and "%D" and
not the slow version for both daily and intraday analysis. I like the "K" values for a quick indications of
overbought and oversold. I have explored different time periods for the oscillator as well as different time
periods for price bars. For the Standard and Poors (S&P) 500, I currently trade 5-minute time bars with a
10-period stochastics oscillator. Unless the market is moving in short cycles with greater amplitude, the
standard 5-period stochastics is too "whippy" on intraday data. This is a recent event stemming from the
fact that larger institutional participation has reduced some of the past futures market volatility.
Chart 1 illustrates a 5-minute bar chart, using the Intra-Day Analyst (IDA) computer software, viewing
the September S&P 500 futures beginning around 1:00 PM on June 28, 1984 and continuing through
June 29. Table 1 (IDA format) provides actual times, prices, and oscillator values necessary to resolve the
patterns. Chart 2 is simply an extension of Chart 1. The "K" line is drawn as solid and "%D" line is
dashed. Immediately you can see that these lines are not nearly as smooth nor as crisp as those illustrated
on page 99-100 in the May 1984 issue of Technical Analysis of Stocks & Commodities magazine.
A trendline established on June 28,1984 is drawn in via the trendline function of Intra-Day Analyst. The
following is a case study discussion of the Stochastics cyclic highs and lows (labeled as points A through
J):
CHART 1:
A: Provides a trendline contact point and is my first reference point.
B: The S&P 500 Futures rally to new highs but, with no divergences thus implying a continued uptrend.
C: The point A lows hold with the %D higher than A. The uptrend is still intact.
C1:A lower opening breaks below the lowest price made at point C by only one tick and then closes at
the 5-minute time bar high, yet %D is rising. I consider this to be illustrative of a "K" line failure. The
rising %D implies further price strength. An aggressive trader would buy this pattern with a stop loss
below the point A price low. The next period retests the low and holds. The market is showing good
support at the 154.90 to 155.00 level.
D "K" reaches 100% and %D peaks at a price of 155.95. No divergence here. This could be a high but
markets have a tendency to take out "even money" levels.
Chart 1:
Chart 2:
Stocks & Commodities V. 3:1 (28-29): Day Trading With Stochastics by John F. Kepka
E This is a "Set-Up" with "K" with a higher low and %D with lower low. The rule suggests to sell the
next rally.
F A new high is made by only one tick with a bearish divergence in the Stochastics. This could be the
high of the move! I would liquidate the long position established at point C1. Traders with deep
pockets full of cash could sell at this point, risking 50 to 75 points since resistance has not been clearly
established. Caution: Note that the uptrend is still intact, according to the trendline.
G This an example of another "Set-Up". The saliva is really starting to flow now. By now there is
significant evidence that the rallies are weaker and the declines are stronger.
H New price highs for the move on declining "K" and "%D" values as well as a small price reversal.
This signal tells me to sell short, I use a 50 to 75 point stop. I consider point H to be an example of a
Hinge in "K".
CHART 2:
H Chart 2 shows that the market did in fact spend 20 to 30 minutes in the 156.20 to 156.30 level with
%D values declining consistently and three thrusts to the upside nicely shown by the "K" line. This is
the main reason why I like the regular "K" line, over the slow version, to show price thrusts in detail.
Resistance is now clearly established at 156.30. Points D, F and H are examples of triple divergence. I
would sell short with stops above 156.30 with the previous short still in effect.
I Market is very weak and the trendline is in danger of being broken.
J This cyclic rally is "flat" and another sell is signaled at a time of 2:28 PM. In the next five minutes, the
price breaks the trendline and collapses 70 points. The market is clearly in a downtrend now. A new
intraday low of 154.75 is set and price closes at 154.85 with ample time to cover shorts.
These Stochastics signals were in no way perfect text book examples and the best sell signal J (resulting
in the biggest drop) occurs around %D of 50% (not 80%). Many times I have observed that a point J type
signal portends pronounced weakness to come shortly. The reverse is true for an uptrend.
I must conclude that the use of Stochastics is very useful but, requires considerable interpretation.
Trading is still just as much "art" as it is "science".