Stocks & Commodities V. 28:6 (22-27): The Forex Gambit by Walter T.

Downs

Game Plan In Accordance

The Forex Gambit

This trading strategy uses logic and analysis from math and from chess and math, I will refer to chess as “C,” and to math
BRIAN TAYLOR

chess. It’s bold, it’s aggressive, and you don’t lose sight of risk. as “M”:

I
by Walter T. Downs C and M both favor economy of time and effort in relation
to what is achieved.
am an advanced chess player, and I have an advanced In this regard, a daily time frame is the best choice. C and
education in esoteric mathematics. Together, these two M both state that the optimal strategy is composed of the
disciplines embody powerful logic and methods of most prolific factors within the general theory and body of
analysis. I have created a system based on the idea of knowledge that governs the system. The general theory of
the gambit, a term borrowed from chess. In its essence, technical analysis is:
it is the sacrifice of material in order to launch a bold
n Market action discounts everything.
attack. The gambit is everything a trading system should be:
audacious, dynamic, and fluid. When it is successful, it reaps n Prices move in trends.
great rewards.
n History will repeat itself.
Here, when I am referring to logic and analytical technique
Copyright (c) Technical Analysis Inc.
Stocks & Commodities V. 28:6 (22-27): The Forex Gambit by Walter T. Downs
FOREX FOCUS

Within the body of knowledge, the most prolific factor that The system
governs trend is the simple moving average (Sma). The most
Buy two lower lows and a closing price that is greater
prolific price used is the closing price. The period used for
n
than the midpoint of the bar and is above the Sma.
the Sma is determined by M and is a value that has statistical
significance. Therefore, the value used is 30. n Place a stop-loss order just below the entry bar.
Within the body of knowledge, the most prolific factors that
govern market entry are the correction (a brief countermove n Sell two higher highs and a closing price that is less
before the trend resumes) and the breakout (price action mov- than the midpoint of the bar and is below the Sma.
ing above or below a defined point). n Place a stop-loss order just above the entry bar.
C and M state that the factor used should be the one with
the least risk and greatest reward potential. In this regard, the n When a bar occurs whose range is greater than the
correction is the right choice. Within the body of knowledge, previous three ranges and whose close is in the top
the most prolific factors of the correction are an action (coun- 10% of the bar, exit a long position on the first closing
termove) and reaction (proof of resumption of the trend). price that is lower than the previous closing price.
Here is a simple definition of a correction: n When a bar occurs whose range is greater than the
n A bullish correction is two lower lows (action) and previous three ranges and whose close is in the bottom
a closing price greater than the midpoint of the bar 10% of the bar, exit a short position on the first closing
(reaction proving resumption of uptrend). price that is higher than the previous closing price.

n A bearish correction is two higher highs (action) n Close any position that has been in the market 16 days.
and a closing price less than the midpoint of the bar n Move stop-loss orders to break even on any trade that
(reaction proving resumption of downtrend). has been in the market two days.

Now, let’s specify our risk point. This is the point at which n Take only one trade at a time. Another trade is not
we will admit that our prediction of future events is wrong. opened until this trade is closed. Often, you will have
C and M dictate that the most prolific option should be multiple trade entry signals in several markets. Take
used. For long positions, this point is just below the low of the the trade with the least risk.
bar on which you entered the trade. For short positions, this n Trade all available currency pairs.
point is just above the high of the bar on which you entered
the trade. According to the logic and analysis of C and M, this is an
Now, you must specify the points at which you will exit efficient system. Figure 1 shows a buy pattern, exit condition
a trade that does not hit your risk point and moves in your met, and long position exit. Figure 2 shows a sell pattern, exit
favor. C and M state that these points are dictated by time, condition met, and short position exit. It is at this point we
and the most likely point at which a bullish or bearish phase will use C and M to create the forex gambit.
is complete. You need to establish a trading account that represents
minimum risk. First of all, open a micro trading account
Taking this into consideration, we come up with the fol- with a reputable forex broker. (The brokerage should have a
lowing exits: deep financial base, and be a member of the Cftc and Nfa,
n When a bar occurs the range of which is greater than the two organizations that look after investor interests in the
the previous three ranges and the close of which is United States.)
in the top 10% of the bar, exit a long position on the Fund the account with $300. Your leverage level should
first closing price that is lower than the previous be set to 100 to 1, which is the maximum leverage allowable
closing price. under the law. Your unit size should be 1,000 units.
What will this mean to you in practical terms? If you open
n When a bar occurs the range of which is greater than a one-lot position, the margin requirement would be $8–16,
the previous three ranges and the close of which is and the value of your position would be $0.08–0.13 per pip
in the bottom 10% of the bar, exit a short position on (also known as a tick).
the first closing price that is higher than the previous According to M, C, and gambit play, we will now seek
closing price. to gain maximum profit in the shortest possible time. Thus,
n Close any position that has been in the market 16 we come to the aspect of bet/position sizing. We will use
days. the logic of M to place optimal bets. The number of lots to
trade is the number that brings the dollar amount risked plus
We now have all the factors we need to build the system. margin closest to but not greater than your trading equity.
Here’s an example:

Copyright (c) Technical Analysis Inc.

 Stocks & Commodities V. 28:6 (22-27): The Forex Gambit by Walter T. Downs
EUR/USD Bid MovS=1.3769
1.5050
1.5000
1.4950
you begin trading, treat $200 of this
1.4900
amount as your reserve, and consider
1.4850 $100 as your trading equity. Any
1.4800 amount of money you accumulate
1.4750 over and above your reserve is treated
as trading equity. So if you had an
1.4700 initial trading equity of $100 and
1.4650 you made $50, you would have a
1.4600 trading equity of $150. This amount
would be used to calculate your bet/
1.4550
position size.
1.4500 Thus, your bet/position size
FXAccucharts.COM

1.4450 varies with your trading equity. If

your trading equity is reduced to an
’09 23 24 25 28 29 30 Oct 02 05 06 07 08 09 12 13 14 15 16 19 20 21 22 23 26 27 28 29 30
amount that does not allow enough
FIGURE 1: A BUY PATTERN. The blue square indicates a completed buy pattern (two lower lows, and a C > midpoint for multiple lots, reduce your posi-
tions to one lot.
of bar and > SMA). The turquoise square indicates that the exit condition has been met (A range > three previous
ranges and a C in top 10% of the bar). The yellow square indicates the trade exit (C < previous C).
If total equity were $205, you
would only have $5.00 above your
Margin = 10.00 reserve. Trade only one lot. If you lose more and move deeper
Risk = 50 (the difference between entry price into your reserve, continue to trade only one lot. Continue to
and your stop-loss order) do this until your trading equity returns to a level above your
One lot = 10 cents per pip $200 reserve level. Once you move back above your reserve
Trading equity = $100 level, you can once again trade multiple lots.
You would buy or sell six lots. As an example, if your total equity were $195, you would
trade one lot. If you then had a winning trade and your total
Margin x 6 = $60 equity went to $250, you would have a trading equity of $50
Six lots x 10 cents = 60 cents per pip x pips at risk and would have enough to apply the bet sizing formula and
(50) = $30 trade multiple lots:
Total = $90 (nearest to but not greater than
the trading equity) Total equity ($250) – reserve ($200) = $50 trading equity
You funded your forex micro account with $300. When
Testing the forex gambit
Now we will test and analyze our
system. We will explore some con-
GBP/USD Bid MovS=1.5721
cepts familiar to C but are not in the
1.6700 general theory and knowledge base
1.6650
1.6600 of accepted techniques for testing and
1.6550 analyzing market systems.
1.6500 C understands a dynamic and
1.6450
1.6400
ever-changing game. It knows that
1.6350 the patterns of the system will have
1.6300 variations that deviate in small ways.
1.6250 The accepted system testing applica-
1.6200
1.6150 tion is to pop the pattern, stops, and
1.6100 take-profit algorithm into a computer
1.6050 and consider the testing results as a
1.6000 valid representation of how well the
1.5950
1.5900 system will operate in the market.
1.5850 To C, this approach is an er-
1.5800 ror in logic. The computer functions
1.5750 in a vacuum. It does not consider
’09 02 03 04 07 08 09 10 11 14 15 16 17 18 21 22 23 24 25 28 29 30 Oct 02 05 06 07 08 09 12 13 14 variations as patterns unfold and
FIGURE 2: A SELL PATTERN. The red square indicates a completed sell pattern (two higher highs, and a C < midpoint
play out, nor does it consider the
of bar and < SMA). The turquoise square indicates that the exit condition has been met (A range > three previous ranges various market states that can be in
and a C in the bottom 10% of the bar). The yellow square indicates the trade exit (C > previous C). effect when the patterns occurred.
Copyright (c) Technical Analysis Inc.
Stocks & Commodities V. 28:6 (22-27): The Forex Gambit by Walter T. Downs
FOREX FOCUS

The computer is not dynamic in its interpretation of events. capable of. It has been my personal experience using highly
It is a static interpretation of a dynamic, fluid environment. systematic discretion (Hsd) that the generated statistics (not
Computer testing of a system often seriously understates counting the bet sizing algorithm that is a feature of the forex
the viability of a system. The cure for this is to walk through gambit) are not out of line with what occurs in real trading. As
your trades manually. Just as a chessmaster alters his game the Hsd trader develops a deeper understanding of variations,
plan in accordance with the exact position on the board, so too the system results just keep getting better.
should the trader alter his game plan as events are unfolding. Now that we have the test results, we should take a look
The trader should retain the patterns theme, but alter his trad- at the logic of the bet/position sizing algorithm. Is it too ag-
ing in dynamic ways in response to the variation he is seeing gressive? Is our gambit unsound?
at that particular moment. To answer that question, I call on M and the Kelly formula†.
Some possible variations to consider when walking through The Kelly formula is defined as:
the trades are:
(Odds x winning percent – losing percent)/Odds
n A windfall profit made in just a few days — consider
exiting. The formula specifies a maximum bet size compared to the
equity of your account. If the percentage of equity that you
A closing price very near the midpoint criteria — consider
bet exceeds the formula, it is considered too aggressive or a
n
entering.
foolish bet.
n A buy or sell pattern that occurs after price has had a Combining all our test results, we get an average winning
parabolic move and is very distant from the Sma — percentage of 67% and average odds (risk to reward) of 3.03:1.
consider standing aside. Applying the Kelly formula, we see that we should risk up to
56% of equity per trade. The most our algorithm ever bets is a
When you are sitting on a large open profit and the exit
maximum of about 40% of equity. We are not exceeding the
n
conditions are almost correct — consider exiting.
Kelly limit. This implies that our betting scheme is smart.
n A buy or sell pattern that is almost correct and is very The other question that springs to mind is, given the test
near the Sma — consider entering. results, what is the risk of ruin, or odds that I will deplete my
$300 if I follow the betting algorithm and account manage-
In accordance with C, walking through the trades also car- ment criteria? The answer from M is that you have a less than
ries a powerful benefit. As you walk through the trades and 1% chance of that happening. See the sidebar on page 26 to
continually study the variations, you will get a feel for how determine how I came to the 1%.
patterns play out in the future. Your grasp of the variations
and powers of intuition will grow stronger as time passes, and Summing up
so too will your success with the system. We have based our system on the most efficient technical
Chess players who practice this method can look at any foundation. In order for us to fail, technical analysis itself
variation of a pattern and pick the right move for the position must fail. The forex gambit is what trading is all about. It is
without the need for deep analysis or calculation. Bearing in a bold, aggressive attempt to gain the maximum profit while
mind the strong analogy between C and trading, it is logical not losing sight of risk.
to assume that the same results will occur. Instead of underestimating systems with computer testing,
You should trade mechanically and remove all subjectivity we choose the Hsd walk-through method, which errs on the
from trading. C, and the tenets that guide gambit play, suggest side of optimism, and we hope that we have made a case for
an alternate approach: our beliefs being closer to reality than the commonly accepted
approaches.
A trader’s style should be one of highly
Hsd will turn you into a real trader, not a computer techni-
systematic discretion (Hsd)
cian! The worst you can do is lose $300, and the profit potential
is, for all intents and purposes, unlimited. So why not give
The trader honors the theme of the pattern but considers the forex gambit a try?
variations as well. Thus, the trader remains fluid and dynamic Take any forex currency pair and perform your own walk-
in an environment that is exactly that. through testing using Hsd and the betting algorithm. Further, it
would be of interest to code the rules and test the system with
System test results a computer to see just how divergent a mechanical approach
In three runs of 90 random trades, Figure 3 shows the variances is in comparison to the Hsd test results.
you can expect with the forex gambit. Two runs scored excellent
profits, earning $38,478.61 and $84,089.81, respectively. The Walter Downs is an independent systems analyst and trader. He
third was a home run, earning $11,423,358.91. The time taken holds a doctorate in esoteric mathematics and is a chessmaster
to complete the 90 trades was 159 trading days for each run. and a member of the International High IQ Society. Currently,
Rather than the underestimated findings of a computer test, he is involved in implementing optimal mathematical forex
our approach takes an optimistic view what the system is systems for fx-mathematics.com.

Copyright (c) Technical Analysis Inc.

Stocks & Commodities V. 28:6 (22-27): The Forex Gambit by Walter T. Downs
FOREX FOCUS

Risk of ruin for the forex gambit

If TheError = 0 Then
Here’s how the risk of ruin (Ror) was determined for the forex
POR = 1 - POR
gambit. In order to determine the “risk of ruin” or “gambler’s
Else
ruin” as it is commonly referred to, the following formula was
POR = “Error!”
used (written in a format that can be plugged into Microsoft
End If
Excel as a spreadsheet macro):
End Function
Function POR(Win, Loss, Pct, Start, Success, Failure)
(where “Win” = avg. win, “Loss” = avg. loss, “Pct” = winning percent,
‘ error checking
“Start” = initial account balance, “Success” = amount that is your
TheError = 1
goal, “Failure” = amount remaining in account that is considered
If Success >= Start And Start >= Failure Then TheError = 0
to be a failure.)
When your account is reduced to $200, you bet only one lot. The
‘ convert data into “coin-flip” approximations
figures that get plugged into the Ror formula are derived from the
Expectation = (Win * Pct) - (Loss * (1 - Pct))
statistics that are in the Figure 3 spreadsheet. So your Ror is:
StdDev = (((Pct * (Win * Win)) + ((1 - Pct) * (Loss * Loss))) ^ 0.5)
P = 0.5 + (Expectation / (2 * StdDev))
Win: $39 (388-pip average x 0.10 (1 lot))
Q=1-P
Loss: $13 (131-pip average x 0.10 (1 lot))
Pct: 67%
‘ find units of capital until success or failure
Start: $200
Account = (Start - Failure) / StdDev
Success: $1,000,000
FinAccount = (Success - Failure) / StdDev
Failure: $0
‘ apply formula (we must do an averaging if P=0.5)
Plugged into the Ror formula, these statistics produce an Ror of
If P <> 0.5 Then
POR = (1 / (1 - ((Q / P) ^ FinAccount))) * (1 - ((Q / P) ^ Account)) 0.000055, notably less than 0.01 (1%).
—W.D.
Else
r1 = 0.50001 / 0.49999
r2 = 0.49999 / 0.50001
por1 = (1 / (1 - (r1 ^ FinAccount))) * (1 - (r1 ^ Account))
por2 = (1 / (1 - (r2 ^ FinAccount))) * (1 - (r2 ^ Account))
POR = (por1 + por2) / 2

End If

Copyright (c) Technical Analysis Inc.

Stocks & Commodities V. 28:6 (22-27): The Forex Gambit by Walter T. Downs
FOREX FOCUS

$100.00 $100.00 $100.00

1 450 $289.00 0 $100 -100 $60.00
2 0 $280.00 600 $340.00 0 $60.00
3 300 $616.00 300 $748.00 0 $60.00
4 -120 $320.32 700 $2,842.40 250 $120.00
5 100 $448.45 -150 $1,136.96 150 $192.00
6 100 $627.83 0 $1,136.96 150 $307.20
7 -150 $439.49 700 $4,320.45 75 $399.36
8 200 $791.08 0 $4,320.45 600 $1,357.82
9 600 $2,689.68 125 $6,480.67 850 $5,974.43
10 0 $2,689.68 0 $6,480.67 -200 $3,584.76
11 200 $4,841.42 0 $6,480.67 -250 $1,786.00
12 300 $10,651.13 -200 $1,296.13 0 $1,786.00
13 0 $10,651.13 600 $4,406.86 -200 $1,071.60
14 -100 $6,390.68 300 $9,695.09 900 $4,929.36
15 175 $10,864.15 0 $9,695.09 0 $4,929.36
16 225 $20,641.89 350 $23,268.20 0 $4,929.36
17 150 $33,072.02 -100 $13,960.92 -125 $2,464.68
18 0 $33,027.02 -100 $8,376.55 -100 $1,478.81
19 -60 $25,100.53 800 $35,181.53 250 $2,957.62
20 -120 $13,052.28 500 $105,544.58 0 $2,957.62
21 400 $33,935.92 0 $105,544.58 -150 $1,626.70
22 500 $101,807.76 -100 $63,326.75 0 $1,626.70
23 0 $101,807.76 700 $240,641.63 0 $1,626.70
24 700 $386,869.50 -125 $120,320.82 150 $2,602.72
25 400 $1,005,860.71 -120 $62,566.83 300 $5,725.98
26 450 $2,816,409.99 0 $62,566.83 0 $5,725.98
27 400 $7,322,665.97 100 $87,593.56 0 $5,725.98
28 400 $19,038,931.51 150 $140,149.69 0 $5,725.98
29 -100 $11,423,358.91 -100 $84,089,81 350 $13,742.36
30 0 $11,423,358.91 0 $84,089,81 450 $38,478.61
Win 18 Win 13 Win 12
Loss 6 Loss 8 Loss 7
Avg win 336 Avg win 455.76 Avg win 372.91
Avg loss 108 Avg loss 124.37 Avg loss 160.71
Win % 75% Win % 62% Win% 63%
R/R ratio 3.11 - 1 R/R ratio 3.66 - 1 R/R ratio 2.32 - 1
159 Trading days

FIGURE 3: SYSTEM TEST RESULTS. The HSD and bet/position sizing algorithm test results
show that the forex gambit is capable of achieving very good profits and, in one case, massive
profits.

S&C

Copyright (c) Technical Analysis Inc.