Risk Management

any successful trader will agree that Risk Management is the key to trading over the long term (and even the short term );

therefore, this thread is dedicated to Risk Management.

those new to trading, do not take this info lightly!

The basic problem with the system as described by VDeluca is the proposal's poor Money Management. The system assumes beginning with $10,000 equity and then each entry enters with one full lot, nominally, $1,000. That represents 10% of equity for each entry. Please pay more attention to trading systems research; do not focus too much on entry and exit subsystems (8 or 9am candle).

No sane professional trader (or experienced amateur) would risk 10% equity per trade!! The generally accepted level of risk is 1 to 2% per entry, and no more than 5% in the market at one time.

The best reference that I have found that discusses a COMPLETE TRADING SYSTEM includes the following major subsystems: what to trade, how much to risk (position sizing), when to buy or sell (signals), stops, and planned exits (I have ignored tactics). The reference for an excellent complete system designed by Richard Dennis, the legendary commodities trader (including foreign exchange), who eventually turned US$400 into US$200 million, and gained additional fame for his Turtle experiment in the 1980’s. You can download his COMPLETE system at www.OriginalTurtles.com. I use the Turtle Trading system for position sizing (including scaling up) and stops. Those were his system's most advanced concepts, both being based on volatility.

I use mathematical indicators for entry and exit (trend and cyclic) that I programmed from concepts I researched on the WWW. I started trading common stocks in 1959 and used technical analysis on an HP-35 calculator in 1978. I have been immersed in researching forex trading since this Jan, 20/7. During that time, I have been doing quite a lot of research regarding the aspect of money management, a.k.a. risk management. It is generally agreed that if your tech. analysis system is at least half-way sensible (via back testing), that your success or utter ruin will be determined by your risk management system and adherence to it.

To recap, trade 1-2% of equity per trade and risk no more than 5% concurrently in the market. (See the Turtle Trading System regarding the relationship between the level of correlation among concurrent trades and position sizing.) In our business, we have witnessed dozens of $10,000 forex accounts that were rendered crippled or destroyed within 2 to 4 weeks when 10% of equity ($1000, full lot) was used per entry. Please note, that a mini account with only $1000 starting equity is just as risky if $100 is bet per trade. So the minimum forex account is more than $5000 if traded with mini lots. [Even $5000 is too small if you have an early run of bad luck.]

Why is 10% too much to risk? If 10% is risked for each trade, you can afford to lose no more then 9 trades because the 10th trade would likely encounter call margin. Anyway, most traders would not keep betting until they lost 90% of their equity. They usually quit after losing 40% to 60%. So, what are the odds of at least 6 losses out of 9 trades? ONLY about 16% or about 1 out of 6! Are you willing to lose 60% of your equity with those odds? Try running a simple BASIC program to simulate 9 trades, each with 50% win/loss probability. Alternatively, try flipping a coin 9 times. If you are not wiped out in 9 tries, keep track of your wins and losses—if you don’t wipe out in 9 trades, the end is nevertheless very near! The odds of 16% is too high to risk my $10,000! [More rigorously, a loss does not necessarily result in full loss of the entry bet, but the general concept still stands, 10% is too much!]

It seems that many participants in the discussion were enamored with the apparent simplicity of the system and the very attractive hoped for result; that was merely the first of the twins of doom--greed and fear.

May the Trend be with you,
Paul Y. Shimada

VDeluca's got it! If you set your stops and let them limit your losses, the risk limit is set by the maximum allowed losses, not by the entry amount. However, there are several aspects that confound this apparently simple approach--namely the difficulty of properly setting stops and balancing them with entry amount and profitability. First, let's focus on stops.

(1) If you set stops too small, you nearly always guarantee a loss.
(2) If you set stops too high, when you're on the wrong side of a trend, you lose more (and maybe too much, such as more than 2% of equity).
(3) If you are in the habit of adjusting your stops so you can "let your position ride just a litter more, because the chart will turn anytime now" most likely you still lack the discipline to succeed in forex (or other financial instruments) trading.

So, if you do not suffer from problem #3, the issue is, "How does one set the stop level?"

A lot has been written on the subject of risk and chance. Much of it by seasoned and successful traders, and even more by mathematically inclined gamblers. I was an IB for a local fx company (I only traded my own money). They had an excellent training program regarding introduction to forex and marketing (getting client/investors). The two areas in which they failed miserably were how to manage money and how to enter/exit positions. The advice they gave in class with great conviction and authority was, "After you make your entry, set you stop at 50 pips." Well, I've always been a good student who trusted the wisdom of instructors, so I followed their advice. I traded H1 charts and I lost nearly $2000 in three days because of problem #1. So as a good student, I have immersed myself in a 5-month intensive research effort. Here is what I've found regarding money (risk) management.

(a) It is preferable to set your stops based on the level of volatility of the chart, which you are trading; one size does not fit all. For example, the volatility could be measured from Average True Range (ATR[20]) as did the Turtles. (As noted in my previous note, I follow the Turtle system for entry size and stops.)

(b) If you find the computation of Turtles "N" too difficult, I would gladly send you an Excel spreadsheet that makes it a simple lookup.

(c) If you want an alternative to the Turtles risk management approach, try this formula:
Stop = (Price) +/- [(Equity) * (%Risk) / (Size of Entry Position) / (Leverage)]
for example, if your entry Price=$1.8321, Equity=$10,000, %Risk=2%, Entry Size=$1,000, Leverage=100:1,
then Stop = $1.8321 +/- 0.0020

That's only +/-20 pips because your entry is 10% of equity. If ATR~50 (typical for many H1 pairs), it is almost certain you would stop out, falling into problem #1.

So what can you do? the Turtles recommend stops at 2N, which is twice ATR. If ATR=50 pips, your Turtle stop would be 100 pips. But +/-100 pips represents 10% of your equity for a full lot, falling into problem #2 (risking more than 2% of equity).

So what can you do? You have two choices. First, reduce your position size to $100 (mini-lot instead of $1000) and keep your stop at 100 pips. You will no longer have either problem 1 or 2, but your trading system better deliver overall better than 2.0 reward/risk otherwise you're doomed anyway.

Second alternative is to change the chart period to reduce ATR. For example, if H1 gives ATR=50, try M30, ATR may be down to about 30. So if you kept your entry size to $1000 and your stop at 2N=2(ATR)=60 pips, your %Risk would be reduced to about 6%. You could change your chart to M15 and %Risk would be reduced further.

However, it would be more prudent to select the first alternative because your risk would be lower (near 1-2%)and you could maintain your trade at your preferred chart period (if you liked H1).

This example illustrates the interaction of volatility (which changes not only among pairs, but also day by day), entry size, %Risk, Equity, and Leverage (remember, leverage cuts two-ways!). Learn these concepts and you will be well served with a long trading life (if you don't take too much risk, have a good entry/exit system [>50% win/loss & >2.0 reward/risk], and have adequate capital).

There are other methods for setting %Risk and Entry Size. A very famous algorithm is known as the Kelly formula. Kelly was a mathematician for Bell Labs and he invented a formula in 1956 to predict the amount of noise on telephone lines. Professional gamblers quickly recognized the applicability of his formula to bet sizing. A simple application of the Kelly algorithm is given by Overholser, Ray (2000), "Money Management: Designing a Money Management Strategy," Technical Analysis of Stocks & Commodities, May 2000, p.38-46. The original Kelly paper (too much math and theory) and many other more useful money management references may be found at http://www.turtletrader.com/default.htm and http://www.turtletrader.com/money.html. While at that Web site, take a look at the Burke, Gibbons (2000), "Managing Your Money" article, which contains other bibliographies on Risk Mgt. Also, Ed Sekota has a refinement on how to define "at risk." All these and more can be found at the above Web pages.

I hope I've whetted your appetites so that you might take more interest in a boring but vital aspect of trading success. Remember, the three rules of trading: (1) Don't lose money, (2) Don't lose money, (3) If you must lose money, lose as little as possible! A broke trader is out of the market and all profit opportunities are wasted on a broke trader!

May the Trend be with you,
Paul Y. Shimada

p.s. To: VDeluca, I realized in your question that you consider 5% equity risk okay, and maybe 2% was too conservative. As my old friend Odd Job (Harold Sakata)used to say (as a Honolulu-based professional wrestler, not in his Goldfinger role), "Patience boy-san, patience!" Two points: (1) many of the top traders in Jack Schwager's "Market Wizards" limited themselves to 2% of equity; (2) try digging out your 8th grade math books and take a look at combinations, permutations, etc.--try computing the probability of losing 50% of your equity if you risk 5% for each trade; those are not odds on which most people would risk their hard-won savings. For example if you lost 10 trades in a row, you would immediately hit your 50% limit. What are the odds of losing 10 out of 15 trades, or 10 out of 20 trades? Maybe you're willing to lose 100%; how long would that take? Of course it gets more complicated if you factor in your win/loss and risk/reward factors. If you have traded long enough to have developed these factors, or if you can guess them, I have an Excel spreadsheet that graphically illustrates the probability of going broke. Hmm, that's a sobering thought.

here are a few places you can read about Risk Management further...

1. Van Tharps TYWTFF: he has a chapter devoted to Position Sizing, the art of limiting your risk by calculating the number of lots you should trade.

2. The Original Turtle Trader: excellent outline on how to setup your stops and position sizing. go to www.originalturtles.org to download the free literature.

One thing i have to respectfully disagree with from these sources is the amount you should risk on each trade. as a swing trader and money manager for others, i risk just 0.7% of the account equity on any given trade, while the turtles used anywhere from 1-4%. Van Tharp recommends 1-2% of total account equity.

pay special attention to how the turtles had the different "levels of risk". this is directly applicable to Forex trading, as MANY trades you make on different currency pairs have a high correlation.

http://members.aon.at/tips/moneyMan.htm
This is an excellent site on Risk/Money Management. Check out the free PDF's to download.

here are a few points worth considering.

1. Many traders have extraordinary luck with demo accounts, partly due to their lack of fear when real money is not involved. I also did extraordinarily well with my demo account and couldn't wait to trade real dollars. I was too bold when I should have been cautious, and too fearful when I should have been bold. I knew about "plan the trade and trade the plan," but I had not yet studied risk management, so I obediently followed my plans and burned a $2K mini account in a week. I certainly do wish you and all traders the best of luck with your live accounts, but sadly, luck usually has little to do with long-term success.

2. Unless you are a very skillful trader, your long-term odds of winning are no better than 50% (that high only if you can control greed and fear). Now consider that the probability of a losing run of 6 in a row has exactly the same probability as a winning run of 6 in a row (both 1/2^6=1/64=1.56%), regardless of previous win/loss history. So your 2 wins in a row were just as likely as 2 losses in a row. So do you think you can extend your winning streak to 5 days? Of course, anything is possible, but what is probable?

3. Your use of 200:1 leverage is extremely risky. Remember that leverage cuts 2-ways--leveraged gains and leveraged losses. 200:1 looks really good when you win, how would 200:1 look with a string of losses?

4. Putting it all together, the coldness of probability, the difficulty of learning to win 50%, the difficulty of achieving risk:reward>=1/2, the ego inflation of a string of wins, and the sting of a string of losses--the greatest difficultly is controlling the rollercoaster of our emotions.

Paul,

Regarding your above post (which I have taken the liberty of copying to this thread as it seems more appropriate here), can you explain how you arrive at a theoretical win probablity of 50%?

Is it as simple as saying that for each position you put on, you either win, or you lose? OR isn't it more complicated than that in that the probability of a winning trade depends upon the particular currency pair and its historical behaviour, coupled with profit and stop loss targets.

For instance, assume that EUR/USD exhibits an average price movement of 76 pips each and every daily session, beginning with the London hours. If you are only going after 10-20 pips profit on each trade, should not your probability of winning be much higher than if you were shooting for 200 pips profit on each trade?

50% is not magic, it's just the mean and median of all trades. A "good" trader should be generally be able to achieve 50% or better--this means he has a well developed system of sense of entry and exit. Inexperienced traders are poor in entry. Additionally, inexperience (a.k.a. greed) often causes traders to overstay their welcome, turning otherwise profitable entries into net losses.

One of my associates has a live account history with better than 80% win/loss. However, his reward/risk ratio is less than 1. Most traders strive for at least 2.0.

The balance of (1) batting average (win/loss) and (2) payoff (reward/risk) determines whether or not you will survive and prosper. What we should strive for are not necessarily targets for highest w/l and r/r, but steady growth of our account equity, the equity curve. Using my Excel "Equity Simulator," we demonstrated that my associates 80% w/l and <1 r/r meets the survivability criteria--he will not go broke and his account will proper, albeit sloooowly due to low r/r.

Yes, if you set profit target only 10-20 pips rather than 200 pips, your w/l should be very good if your enty/exit techniques are sound. However, using 10-20 as profit target, what is your risk? If you set your stop loss at 20 pips with a daily range of 76, you would most likely stop out often, thereby decreasing your w/l ratio. However if you set a reasonable stop at 100-150 (the stop level depends on your trading style and volatility) and maintained your 10-20 pip profit target, your w/l should be fairly high.

However, with profit target~10-20 and stop~100, your reward/risk=20/100=0.2 (if you could win 100% of your entries). So w/l~80% and risk/reward<=0.2, your equity curve would be very choppy and my Equity Simulator shows that you would be broke after about 100 to 200 trades. Hmm.

As I noted in another post on setting stops, this is not a trivial topic. Your stop also measures what percent of equity you risk, which should be about 1% per position.

May the Trend be with you,
Paul Y. Shimada

I wrote the following for posting on the Yahoo MetaTrader forum and then remembered that there was a forum focused on risk management. There's always more to learn!

Subject: Simple statistical assessment among alternative back test strategies

1. For each back test alternative, compute the statistical "expectation."

2. Compare the expectation among alternatives.

3. Alternatives with negative expectation are expected to lose money.

4. Consider only alternatives with positive expectation.

5. If all other considerations are equivalent (such as capital required, attention and effort required, etc.), the alternative with the highest positive expectation would usually be selected.

6. The formula for computing expectation is as follows.
E = [1 + (W / L)] x P - 1
Where
E = expectation
W = average winning amount for all winning trades
L = average losing amount for all losing trades
P = probability of winning, same as percent of winning trades

6a. The following example uses this equation based on historical track record (a back test simulation could be used). The probability of winning was 63 percent and the average winning trade was $454 and the average losing trade was $458, the mathematical expectation is:
[1 + (W / L)] x P - 1 =
[1 + (454 / 458] x 0.63 - 1 =
1.99 x 0.63 - 1 =
0.2537

6b. Compare this with the strategy that has the following statistics.
Average win = $2,025
Average loss = $1,235
Percent profitable = 0.52
(1 + 1.64) x 0.52 - 1 =
1.37 - 1 =
0.37

6c. Case b has slightly higher expectation than case a, and thus, case b may be preferable.

7. Expectation, a mathematical outcome is not predictive in nature and can only be used to gauge the strength of a system's past results. This is, in any case, the only use for historical statistics and back testing.

8. Reference: JONES, Ryan, 1999. The Trading Game, Colorado Springs, Colorado, March 1999, published on the Internet as an Adobe PDF file, 115 p., Web site unknown.

9. So far, this is the most comprehensive, useful, and truly interesting material on money management (usually a dull topic, preferable to counting sheep) that I have found. This book is not about how to make winning trades. It focuses on how to win more and keep more of your winnings--long-term net profit!

Previously, my money management system was to restrict each trade to no more than 1% to 3% of equity. My strategy was focused on survival because too many of my associates went broke risking AT LEAST 10% on each trade. For 10% entries, I computed a 17% failure rate for 7 of 9 trades; too risky for real money! I adopted the 1% concept from Turtle Trading, but I had not considered using statistics to Balance Survival with Profitability. Aha!

The original source PDF file was published two pages per sheet and was too small for me to comfortably read. I am in the process of converting the PDF file to MS Word RTF (317 pages and growing). I would gladly send copies to anyone who is interested after I have finished the laborious conversion cleanup (the numbers and tables convert poorly or not at all). Unfortunately, I do not expect this soon because it is not among my high priorities.

2004.0728, Paul Y. Shimada, PaulYShimada<at>Yahoo.com

10. Teaser: "This attitude is also why many people are getting involved with the commodity and futures industry. Trading can be a powerful endeavor. On the other hand, it can also be financially crippling. Trading is a game of risk versus reward. It is also a game that is not forgiving of players who come in without learning the rules. For those with the "get rich quick" or 'gotta have it now' mentality, failure is all but certain.

The failure rate of those who attempt to trade in the leveraged markets arena is somewhere around 90 percent. As far as I can tell, this means that 90 percent of those who begin trading stop showing a net loss. I have also been told that at any given time 90 percent of the open accounts show losses while only 10 percent of the accounts show profits. These statistics illustrate that getting rich quick in these markets is highly improbable. To make serious money in this environment, traders must manage their money. Unless sheer luck intervenes, no one will make a fortune in leveraged markets without proper money management strategy. This is the basis of this book"

11. Teaser: " Take a coin and flip it in the air 100 times. Each time the coin lands heads up, you win two dollars. Each time the coin lands tails up, you lose only one dollar. Provided that the coin lands heads up 50 percent of the time and tails up the other 50 percent of the time and you only bet one dollar on each flip of the coin, after 100 flips, you should have won a total of $50.

100 flips
50 flips land heads up. 50 x $2 = $100
50 flips land tails up. 50 x ($1) = ($50)
$100 + ($50) = $50

(Note: This is a fictitious game. I have had some traders call me and tell me that this doesn't simulate real-time trading. My response is that it is not meant to simulate real trading, only to show the power and demise of money management.)

Obviously, this is an ideal betting situation. Since we can spot the profitable opportunities here (being the astute traders that we are), we are not going to bet just one dollar on each flip of the coin. Instead, we have a $100 account to bet in this game. There are many possible ways to bet the scenario. However, you must choose one of the following four options:

A. Bet 10% of the total account on each flip of the trade.
B. Bet 25% of the total account on each flip of the trade.
C. Bet 40% of the total account on each flip of the trade.
D. Bet 51% of the total account on each flip of the trade.

These are the four options. If you choose A, you will multiply the account balance by 10 percent and bet that amount on the next flip of the coin. You will then take the total amount won or lost plus the original amount bet with, place them back into the account and multiply the total by 10 percent again and bet with that amount. Therefore, starting with $100 and multiplying it by 10 percent gives you $10 to bet with on the next flip. If that flip is a winner, you win $2 for every $1 you bet with. Since you bet with $10, you win a total of $20 on the first flip ($10 x $2 = $20). Take the $20 and place it back into the account and you now have $120. Multiply this by 10 percent and you will bet $12 on the next flip. If the next flip is a loser, you will lose only $12 which will bring the account down to $108. You get the picture. Do the same if you choose B, C, or D.

The results are as follows:

A. After 100 flips, $100 turned into $4,700.
B. After 100 flips, $100 turned into $36,100.
C. After 100 flips, $100 turned into $4,700.
D. After 100 flips, $100 dwindled to only $31.

The whys and hows of this illustration will be dealt with later in the book. For now, I want to point out two critical facts about money management. First, it can turn a relatively mediocre trading situation into a dynamic moneymaker. For a trader who staked a flat $10 on every trade without increasing the size of the bet, the net value of the account would have only been at $600. However, increasing and decreasing the amount of each bet increased the return by 683 percent. If a trader would have bet a flat $25 on each flip, the net value of the account would have ended at $1,350. By increasing the amount bet as the account grew, the return was increased by 2,788 percent. If the trader were to bet a flat $40 on each flip, after suffering two losses in a row, the trader would be unable to continue. Therefore by decreasing the amount risked on each flip, the trader was able to stay in the game.

Second, risking too much on each trade can also turn a winning situation into a losing scenario. Even though the trader would never totally deplete the account (theoretically), the decrease would amount to a 79 percent loss after 100 flips.

This illustration shows that improper money management can turn a winning situation into a losing situation. However, no amount of money management will mathematically turn a losing situation into a winning situation."

Paul,
great post.

I also have the 2 pages per 8 1/2" x 11" page version and read it some time ago so my memory of the details is a bit foggy (I wish I had taken a few notes). Off the top of your head or with a quick look to the book, could you refer me to the pages or perhaps explain or both, the differences between fixed percentage and fixed ratio? I know one very successful trader who swears by Ryan Jones' fixed ratio method. I want to try it instead of fixed percentage that I use now.

Also, when you finish your editing, let us know as I would like a copy thanks. Too bad you can't recreate the table and diagrams.

Pipster

Paul,
the link above that you provided for original turtles is parked. Do you know where else I could get a copy of the "complete system"?
Pipster

pipster, i just uploaded the Turtle Rules for you..

http://www.forexfactory.com/turtlerules.pdf

this is an excellent read. almost 30 years later successful trader are still playing by these general rules. the only thing that my research does not agree with is their method for scaling down the trades when a new account dips negative. i only scale the risk up, never down, because i would never trade entry/exit points that have a negative expectancy.

Thanks Merlin

Merlin, could you post or email me with the risk management Excel spreadsheet you mentioned in your post!

Thanks
Jd

Long live the turtle, thanks Merlin.

sorry my english

ya MM is important but as i see it alot of people bailing in the market risking 50 pips and going for lets say 100 well basicly its a good strategy. but to go to the point is that as i see it u need profit on ur trade and account before we sucesfully can trade with moneymanagement then its pure magic here wee need a paper,pencil and common sence and diciplin.. a quick example is we are in a winning trade 1 lot yes many people add to that position here we go another lot now were in 2 lots why not only 1/2 extra lot much easier maybe just for me, in order to find a reasonable place from where to get profit enough from first position to cover an eventually loss from the added 1/2 position and still out with profit

revoke

Here is an excellent excel file for determining position size. It uses the DDE feed from your broker software (if you have one). Metatrader DOES so it works fine!!

I have included the original which works for normally named pairs (EURUSD) as well as one that works for Interbank FX and their abnormal, micro-name pairs (EURUSDm).

Enjoy,

D.

Very nice post! Still digesting these stuff. It looks simple, but not at all.

After a brief search on FF on the subject of risk management, this 12-year-old thread seemed like the only thread that touches base on the Risk Management (in my opinion) part of trading.


So instead of making a new thread, I'd like to pick up this baton where Mr. Merlin left it off in 2004, almost 12 years after. I might not be a great trader yet, but that shouldn't be the reason stop me from contributing my thoughts on this key subject.

Mr. Merlin, if you are still here, thank you for letting me having this opportunity to continue where you left off. I'd like to believe you laid this foundation for me 12 years ago. I wish you the best wherever you are and whatever you are doing.

One thread I want to visit and contribute more ofter is this thread. I made a statement to come back to this thread in 2016, but i next had time to do it.
And here I am in 2017, Let's get right into it.

You out there, please feel free to join me, too.