Money Management 6

back to

Money Management 5

Money Management:

It's something that most people spend very little time on. The most important aspect of trading is not sexy, exciting, or easy to describe, so who would want to spend time on it?
Who cares if I have 8 losers in a row for $500.00 each? It might only take a few trades to make back my $4,000 and more.....

I would suggest a 1% parameter per trade. If you have a winning method, the only way you can lose is by running out of money. Why take the chance? Never risk more than 20% of your account on your total current trades (cumulative risk, remember, all of the trades you are in, can and at some time (!), will be losers)

Trade long term to cut down on the costs of slippage and commission. Roughly 20% of an average account is spent on those costs per year.

If you want to be more advanced, I would devise a sliding scale of initial risk that was quite small when the original starting capital is at risk and increases if one has profits over a self-chosen Mendosa line of acceptable returns. (ie At the beginning of the year or when down for the year while original capital is at risk, the risk would be .5%. Profits between 1-10% would be risked at a 1% rate. Between 10-20% 2% etc. The numbers used are
for example only.)

In a long term trading situation where the market often gets away from the stop creating unenviable situations, I suggest peeling off contracts, but never below 1, to smooth volatility. (i.e. $100,000 account. Buy 5 Silver at 5.00 on the entry signal. Original stop is 496, 1% risk. [¢470,0 - ¢471,0 = $50] At the close several weeks later the price is 600, but the stop has only moved up to 580. The account is now worth $125,000, and the risk, close to stop, is $22,500 or 18.0%. My 'peel off level' is 3%, this is higher than initial risk because it is dealing with open profits. I can risk $3,750. The risk per contract is $1,000; therefore, I can only have 3 contracts. I cover two.)
This will smooth volatility without adversely affecting returns in the long run.
It will hurt returns in massively trending markets, which number one or two a year, but it will help returns on an 'average' move. It will cut the volatility by over 50%.

By combining the sliding scale of initial risk and a sliding peel off %, both per trade and depending on the initial risk, one will see that money management has much more to do with returns than entry exit decisions. However, the best money management in the world will not work if one's methods do not have an edge over the market.
Scot Billington scot.billington@nashville.com

Robert Buran Company bobburan@usa.net

Markets must move in such a manner so as to Frustrate, undermine and defeat the best interests of the majority of market players. According to this rule the majority cannot make money in the marketplace. The markets will totally ignore fundamentals if the majority of the market players act on those fundamentals.

Price movement is predominantly random, i.e. not 100% random, but predominantly random. The "secret" to making money in the market is locating that small portion of market behavior which is not random and exploiting it. I do not feel that conventional technical analysis is of any value in doing this. The problem with technical analysis is that it will present the illusion of uncovering hidden relationships between price behavior and various indicators.
I would submit that all such indicators are as random as the price behavior they attempt to predict and that all profits and losses realized from trading such indicators will be randomly distributed.

First law of price movement is:
If prices move up there is greater probability they will move higher rather than lower.
If prices move down there is greater probability that they will move lower rather than higher.

Second law of price movement is:
If price goes up you must buy the market long and if price goes down you must sell the market short.

The problem is that when we look at the chart we want to buy low and sell high. You cannot, however, buy bottoms and sell tops. You can only follow a trend which has already been established through price movement in the same direction as the position you are taking.

You must understand that when you elect to buy long a market when price rises you are in effect buying the market at the worst possible price at the time of your entry. It's not going to feel good and it's not going to look good on the charts. But by "buying high" you are probably going to be placing yourself on the minority side of the market and therefore assuring yourself of profits.

Stop and Reverse method:

The stop and reverse method involves utilizing some kind of indicator or a price based on an indicator. If the system is long one contract and the market comes back and the price is hit the system sells 2 contracts and reverses to a short position and so on. Of all possible trading strategies I have found this to be the least profitable and grossly inefficient with respect to the use of margin. I will discuss margin efficiency in greater detail later but for now it need only be said that systems that are in the market all the time tie up your margin needlessly. Markets tend to move sideways about 85% of the time and consequently these systems will have your margin tied up doing absolutely nothing for at least 85% of the time. These systems can also whipsaw you to death while moving sideways.

Trailing Stop:

The second most common way mechanical systems take profits is through the use of a "trailing stop." The idea behind a trailing stop is that it allows you to "let your profits run" while at the same time "locking in" any profits you may have already made. My experience with system design and trailing stops has been that the trailing stop is at best a mediocre method of exiting a profitable position. The problem is that if the trailing stop is too tight it results in your having your stop tagged right before the start of a big move. Conversely if your stop is too deep it results in many small profits going to large losses. The other problem I have with trailing stops is more theoretical. With a trailing stop you are trying to take profits only after the market has turned against you. Frequently you are forced to sell out your long position when many others are trying to sell too. You are then moving with the crowd and this is almost inevitably going to cause you excessive losses.

Therefore the rule I have developed with regard to taking profits is:
*** You should try to take profits only when the market is moving strongly in your favor.

This is much more consistent with my contrary philosophy of trading. If you are long a market and the market takes off like a space ship you should sell. By doing so you put yourself on the minority side of the market selling to the majority of panicked buyers. That is how you make money in this game.

Bob Buran:

What should you do, however, if your position starts out bad, get worse and then threatens an uncontrolled hemorrhage of your account equity?
Unfortunately this happens with about 15 or 20% of our trades and our ability to keep these losses within a normal distribution pattern is what makes or breaks us as traders. This is a particularly critical issue if you are using Systems without stops on day of entry. Let me
assure you that you are utterly mad if you place no stop on day of entry. The day you enter a trade is the time you are most likely to be gored.
I originally experimented with using a money management stop strategy on my day of entry. If I were trading three coffee contracts and figured $2500 was as much as I wanted to loose I would put a stop in about 220 points away from where I got in (1 Point = ¢166 to ¢167 = $375 = ¢37500
$2500/3 = $833/Contract ¢250000/3 = ¢83333,3/Contract
¢83333,3/$375 = 222 points

The problem was that on some days 220 points was far more than I should be risking and others 220 points wasn't enough. The market didn't really care about what Bob Buran was willing to risk.

Out of this frustration I developed a simple strategy that probably works better than anything I ever developed. If you are sick of always having your stops run, this simple strategy is going to be a big help. If you got into a trade based on a longer time frame such as a time
frame based on daily data you need to develop a stop loss strategy that is based on a shorter time frame. For starters you should kick up an intra-day bar chart on your on line computer screen and set the bars to something like 3 to 10 minutes. If you are following our trading rules you are going to buy when the market goes up. This upward movement should create some kind of upward wave on the intra-day chart. You should measure this wave from its top to its bottom and if you are long the market you should place your stop at the point that represents a 75% retracement of that wave. If you are short the market you simply
reverse the process. Hence my rule for placing your protective stop is:

Place your protective stop at a point that represents a 75% retracement (5/8 or 6/8) of the wave/move that got you in.

Here again you see why the "buying high" strategy doesn't sell systems. Buying point B (which is the high) looks like a terrible place to enter this market. Why not sell at point B? Or if we have to go long why didn't we buy at point L (which is the low)? Don't despair. Because you feel that way others will feel that way also and so they, the majority of market players, won't buy because it's too scary. The market in the best tradition of the "Rule of the Screw" will sense this hesitation by the timid majority and move much higher. That will encourage the timid majority and they will then jump into this market in a buying frenzy. At that time you will calmly sell your positions back to the frenzied majority and take your profits.

The point I'm making is that when you first get into these trades they seldom look good and you need to use the 75% retracement rule to place a stop so as to give yourself some peace of mind. If you go back and look at Figure 2 you can see how this stop was calculated. I measure from point L (low) to point H (high) and take 75% of that and subtract that from point H to determine the stop which is equivalent to the price shown at point SS (sell stop).

If you are an Elliot Wave purist you may notice that there are other smaller waves in figure 1. Try to keep it simple and try not to miss seeing the forest for the trees. I'm not an Elliot Wave purist and what I do with a 3 to 10 minute chart is to measure from the highest high after your buy point has been hit to the lowest low on the screen.
Usually that is going to be the lowest low in the last day or two. That's what I mean by "the wave that got you in."

The Fibonacci Connection

Some of you sharper readers may at this point notice that maybe I might really just be playing around with Fibonacci ratios. Indeed what we are really saying when we elect to place a stop at "75% retracement of the wave that got you in" is that if the market fails to be supported at the 5/8 or .618 Fibonacci retracement point, it becomes a "Fibonacci failure," a trend reversal and we need to get out of the way of a collapsing market. Believe me you are going to be very happy to be out of the market if these stops are hit and it will be very unusual for the market to "tag" these stops and then move higher. This is the most effective stop loss strategy we use.

Some of you technical analysts may at this point feel somewhat indicated. Here I am telling you first that technical analysis is a lot of baloney and then I turn right around and start using Fibonacci ratios for stop placement. Of course the ratio .618 wasn't invented by a technical analyst. It was known to ancient Greek and Egyptian mathematicians as the Golden Ratio or the Golden Mean and was used in the construction of the Parthenon and the Great Pyramid of Gizeh. Then again maybe I just like Leonardo Fibonacci because he never was a floor trader and never tried to sell me a seminar or a $3,000 piece of trading software. In any case I confess to having almost obsessive fascination with Fibonacci expansion and retracement analysis.
Nevertheless I don't think that at this time I have developed fully all my ideas regarding this. I am, however, giving you the most significant part of my work with these ratios.

Summary and Conclusions

These "laws" (If Prices move up there is greater probability they will move higher rather than lower; If prices move down there is greater probability that they will move lower rather than higher) are permanent, will not break down and cannot change in the future.

Once we have entered a trade based on these rules we will reject traditional "stop and reverse" and "trailing stop" strategies of exiting our trade. Instead we will:

Take profits only when the market is moving strongly in our favor and place our protective stop at 75% retracement of the wave that got us in.

Some of you may at this point be ready to reject these market theories as being far too simple to be useful. Before you toss this manual in the trash, however, I want you to go back to the appendix and look at my equity curve for the past seven years. Look at the summary of my monthly profits from January 1, 1989 when I became fully automated, through June, 1991. Look at the consistent income and small drawdowns.
How many gurus do you know who have included seven years of real-time trading records along with the materials they are selling?

I learned these rules in the marketplace and while attending the "School of Hard Knocks." On the surface they may seem simple, but implementing them in the marketplace is a more complicated process. Later in this course I will show you how you can consistently gain an
edge on the markets and automate a trading system using these same simple rules. Using these strategies you need not fear that these basic rules will break down or stop working. They can't stop working anymore than Newton's Law of Gravity can stop working. I believe if we stop looking at all those wiggly lines, charts and complicated formulas and concentrate instead on simple up and down price movement we can beat the pants off the big boys. Call this back-to-the-basics trading or call it anything you like. I call it financial security.

Design a system that puts you in an average of 2 markets every day and never more than 5 markets. The drawdowns were extraordinarily small and the equity curve was amazingly smooth. Based on my testing I figured I could make about $100,000 per year with a $60,000 account.

Random Distribution of Profits and Losses, the Professionals' Nemesis

A system cannot know what the market is doing after entry. Your trade plan can. That is your edge. It is not second-guessing but intelligence-gathering upon entry. Systems may be giving you a signal again and again. Does this mean to add at every signal? Your trade plan must address that. I have liked the three add-on points. Use your own ideas.

Don't draw any conclusions about a system (or indicator) on the basis of isolated examples. The only way you can determine if a system has any value is by testing it (without benefit of hindsight) over an extended time period for a broad range of markets.

Exits:

1. Exit at a target, i.e. a retest
2. Exit at a target, i.e. an expansion
3. Exit at a target, a fixed profit objective in dollars
4. Exit at a target, an objective chart point
5. Exit at a target, a subjective chart point
6. Exit at a time interval, 4 days
7. Exit at a combination profit time, first profitable open, 2nd profitable close.
8. Don't exit but just reverse
9. Exit on a range expansion
10. trail a stop off the low
11. trail a stop off the high
12. parabolic exit
13. Exit on close
14. exit if the day's close is less then 66% of the daily range in the direction you are trading

On backtesting, find the maximum profit on each entry before a reverse signal is triggered. Then use something like a standard deviation of maximum profits to find a probable profit range. Take profits on partial position when that profit target is hit, then run a trailing stop.
On stock system trades my first target is 2%, then a trailing stop of 62% of maximium move after 3 days. The system takes advantage of intermediate term swing moves (days to weeks), hopefully catching a few trends. I almost always get knocked out of trend moves, so am designing ways to re-enter the trend after being stopped out. The simplest is to re-enter
when the high (for long) or low (for short) bar is taken out on close. Then I need a stop system on that re-entry.
-
A.J. Carisse carisse@brunnet.net

> This means taking smaller profits and not letting any get away from you,
> it also does not mean trailing a stop because if you do, you will get an even smaller profit.

Not sure about that. This will get you out early too much. One of the difficult things to properly realize is that the performance of an instrument does not bear any relationship to your specific entry. By concentrating on its current characteristics in relation to your entry
point, your exits will suffer significantly. You must let the instrument tell you when it is time to exit. Simply put, this means - the point when you can calculate that your capital would be better off in cash or in something else. Patterns, ranges, and TA can give us valuable input to where this point is, but one thing's for sure, it has nothing to do with how little or much you've made so far on the trade. As far as moving stops go, at least this is relevant, but they are still somewhat arbitrary, and at best are an over generalization (i.e. always exiting on an X retracement without regard to the peculiarities between patterns).

This is a complex matter, and one's approach will vary according to what is traded and one's preferred style. However, there are a couple of principles that would apply universally. In all cases one should seek price confirmation for one's decisions. Ultimately, price patterns are king, and it is wise to at least wait for a reasonable indication that the trading is starting to go the wrong way. As well, while we need to spend a great deal of effort on trying to formulate efficient exit strategies, all this is wasted if we don't stick with whatever rules we have formulated. This isn't always easy in the heat of a trade, but I've found that more often than not, sticking to our well thought out plans will prove to be more profitable.

> So, what I am thinking is to exit when the trade is moving in your favor
> and as long as you made money, be happy.

You'll be happier, though, if you can develop more efficient exits. In terms of risk, i've found that the old adage of giving your profitable trades a little more leeway (not too much of course) makes a lot of sense. It's important, though, that you consider the option of re-entry into your calculations, which is a tool that can allow you to tighten up your exits a little more while still capturing a large part of the bigger moves.

> I would like to hear from experienced traders on this philosophy. I have
> had many times where I have had 500 profit into a trade and did not take
> it only to have the trade fail to my stop. I am a one lot trader so it
> is hard to scale in and out.

I never let profitable trades go south. You need to tailor you exits a little more to performance, as for sure there isn't any way that one should go from a good profit to a good size loss. Having separate objectives is usually the main culprit here, as is setting your exit criteria too loosely. Ideally, an exit strategy should be totally removed from your entry, and just focus on the thing being traded. As simple example, let's say that you plan to exit whenever a set of MA's cross. So - either they cross, or you're still in the trade - regardless of your P/L at any point (which must be kept separate). This doesn't mean that objectives shouldn't be considered - but in terms of the instrument's performance, and not yours. For instance, if you can sense that a meaningful retracement is at hand, on the basis of its recent range of similar moves, and the price pattern suggests that it is underway, you could consider exiting, but this still has nothing to do with where you entered.

> I don't let 500 get away from me anymore, but then I am not letting
> profits run either. I suppose a solution would be to build up the acct
> with one lots and later use three lots to scale out and leave one on for the big pull.

I'm not a big fan of this approach, simply because the odds are either with the play or they aren't. Therefore, there is a right and wrong answer in terms of whether to hold or not, and the task is to try to determine which is the case. Whenever you focus on P/L, you will tend to be overcautious, and believe me, I've learned this the hard way. It is much better to focus
exclusively on the present (the short term odds), while putting the past behind you (where you entered, and thus your current P/L), and letting the future take care of itself.

Regards, A.J.

Bill Shumake bshumake@our-town.com

The first is to always trade a fixed unit of contracts per money invested. In other words a person may decide to trade one contract for every $50,000 of capital, in this way risk-of-ruin is always kept constant.

A second approach was suggested in which the number of contracts traded is increased after the second or third consecutive loss.
The idea being that if a trader has a system that say 60 % correct on average, after two consecutive losses the probability that the next trade will be a winner is about 84 % and if three consecutive losses occur the probability is 93.6 % that the next trade will produce a winning trade.

A third approach, which is simply a conservative variation of the second, is to only take trades after two or three consecutive losses occur. This is a very interesting approach which warrants investigation. It is based on the idea that all freely traded markets go through bases of accumulation/distribution and then an explosive trending move followed by another basing period followed by another explosive move. The money is made, of course, in the explosive moves. If you could isolate the explosive moves you would have the perfect system. This approach attempts to isolate these moves by waiting for two or three consecutive losses, which typically occur during a basing period, before a trade is taken. When a trade is taken, the odds of it catching a trending phase is increased.

With regards to strict money management ideas such as only risking 1% or 2% of total equity on any given trade, the reason I think you find so little written on the subject is that there is not much to write about. What I mean by that is that a trader has very little information to base money management decisions on. For example, a great deal has been written about money management with regards to playing black-jack. One reason is that a black-jack player has information about the odds of winning a given hand, information about exactly how much is at risk and information about exactly how much stands to be won on any given hand. This information allows the black-jack player to use money management systems that allow him to vary the size of the bet in order to maximize his ultimate winnings. Many people have tried to adopt similar strategies for trading. Unfortunately as traders we only have information regarding two areas; what our approximate odds are of winning a given trade, and how much we stand to lose if we do not win. Unlike the black-jack player, we have absolutely no information regarding how much we might win on any given trade, should it turn out to be a winner. (We have an idea of what the average winning trade has been in the past, but no information on any one specific trade in the future.) It is because of the lack of information on this one critical area, "how much will be won on this trade," that most money management systems deal only with putting a certain percentage of capital at risk. However, I believe this is entirely adequate, because ultimately money management exists to do one thing; to make sure you have enough money to stay in the game and continue to trade until winnings offset losses and a profit is realized. If you have a system that is 60% accurate and limit your loss per trade to approximately 1% of total equity your probability of success approaches 100%.

Something else you should consider when thinking about money management and the idea of only risking 1% per trade is the psychological effect. As you no doubt know, one of the key elements to successful trading is being disciplined in applying the rules of your system over time. Those who can consistently apply their system usually win and those who cannot be consistent ultimately lose. Only risking 1% on any given trade means I can be wrong more than 50 times in a row and still have over half my original money. Place this in the context of a system that is 60% correct, where one might routinely be wrong only 4 or 5 time in a row and you can see that it makes trading a much less stressful endeavor and insures that it will never be difficult to pull the trigger on a trade. In other words it makes it easy to be disciplined and consistent with your trading strategy, which in turn guarantees your success. To give a real life example, I read the other day that Linda Raschke typically trades one contract per $100,000. When I read this I thought back to when I was a young boy and my father would attempt to trade futures on a $3000 account (he was never successful by the way). Can you imagine how much easier it must be for Linda Raschke to pull the trigger on the next trade after experiencing a loss than it was for my father when he had just experienced a loss. It is this psychological factor of money management that I personally believe is the most important. Remember, trading is only fun and worth the profits if you can sleep well at night. This is why I believe you see so much about risking only 1 or 2 percent per trade...it is very simple, and while it may not be the path to maximizing gains, it does insure that the trader will always be able take the next trade and apply his system with discipline and consistency. In short, risking only 1% per trade may not make financial sense but it does make Peace-of-mind sense, and in this business, peace of mind is itself a valuable commodity.

Jack Schwager: New Market Wizards:

- nothing else but Volatility can be measured

- when a bigger than daily average move occurs, there is a 55% chance of being followed by a similar directional move on the following day (probability edge is not sufficient after allowing for slippage & commissions).

- 1 coin: 55% chance of landing on heads (i.e. odds of getting heads of a single coin is 55%)
- 9 coins: 62% chance of getting more heads than tails
- 99 coins: 75% chance of getting more heads than tails (binomial probability distribution)

- Fund:
Average Daily Volatility = 1% (1% change per day)
Annual Return = 20% approximately 54% of the days were up (135 of 250 days), 46% of the days were down (115 of 250 days)
Annual Return = -20% approximately 46% of the days were up, 54% of the days were down

- the worse slippage, the better the trade
- Fundamentals = funny mentals (Ed Seykota)

Coin Toss:
Ted Juszczak tedj@net-link.net

In a system that is correct 50% of the time (which is typical of most trading indicators and flipping a coin too) strings of losers and winners will approach 15 in a row. In other words, if a coin is flipped 1000 times, probably (key word) at least once, heads will occur 15 times in a row, tails 15 times in a row also.
There will be even greater occurances of 10 and 11 heads in a row, and numerous runs of 4 and 5. Very few indicators are better than a coin toss.
The same can be true of winning and losing trades. Those strings of losers will occur even more frequently in a system that has 40% winners. A winning system only needs bigger wins by a certain percentage. The probability of winning times the average win minus the probability of losing times the average loss just needs to be greater than zero to generate profits over time. A casino wins about 55% of the time - they wired it into the rules of the game. With only a 5% edge, they can pay all of those employees, give away free drinks and meals, build lavish palaces and subsidize travel. The casinos in New Jersey alone, paid taxes on $6 billion from the take on slots alone in recent years. Your broker and and his colleages in the pits get about 5% slippage and commission. If you use just about any indicator it's going to be "right" about 50% of the time, leaving you with a 45% system. You just have to make sure your average win is greater than your average loser, and not quit just because you started off with that string of 15 losers in a row.

Money management and Trading business
Gwenn Ael Gautier Gw.Gautier@wanadoo.fr

1 - Trading constant size is riskier than increasing size. Indeed, as traders you have fixed costs, and you have to face unexpected events in trading and or life. Let's say you have
- a $100.000 account to live off
- a 100% return yearly
- $80.000 in living, tax and trading expenses yearly

If you are trading constant size, you'll move only very slowly over the years out of your undercapitalized status.
If you are trading with a view to increase along the way, it may still take some time, but you will indeed take off.

Now if in two years you have to face a $100.000 instant loss due to an accident, illness or something, which situation would you rather be in? The prospect of higher risk due to increased size, is also a prospect of lower exposure to unforeseen bankruptcy. I vote for the first.

2 - Increasing size is to be tested just as a system is. Sampling, back testing, forward testing etc. Check all ratios. Is your system displaying typical behaviors? Do winning streaks follow losing streaks, or are outcomes randomly distributed? What are your chances of having awiner after a winner? Two winners? etc.
What are your chances of having a winning streak, after a losing one? A positive return after a negative one? Is their seasonality? Depending on the above, you may increase in %, in steps or according to immediate results, period etc.

3 - Consider trading as a losing game, hence play very conservative as long as your not in a winning position, find ways to be agressive when you are. May be add along the way, or play much bigger in some circumstances.

4 - ABOVE ALL, BE ABSOLUTELY CONSISTENT. With money management, more than anything else, lack of consistency will totally ruin your best laid plans.

5 - Take out profits on a consistent basis, in a consistent way. Either you pull winnings, or you pull revenues, but decide once and for all and TEST, TEST TEST everything. These steps should account foreway over 50% of the time you spent on testing your system.

6 - Same for breaks, holidays etc.

7 - Test for capital accumulation. Your plan should allow self insurance.
Remember, anything can happen, and you want to still be around tomorrow.
Plan for the B scenario, your retirement, etc...

Then you may also look at system diversification, market diversification:

8 - I only trade individually working systems, which also work on a combined basis. Last thing I want is having all drawing down together all the time.
For some time I traded a group of 4 uncorrelated systems, two of which were underperformers, but very uncorrelated. Guess what, they "max drew down" all together, tripling a 15 year combined max drawdown (or over 20.000 combined trades tested).

9 - Know when to stop something that is not working. Just as for a single trade, your activity must have a stop, if things are not right (see 8 above).

10 - Market or product diversification: Again, each must stand its own test, and in combination. For me option writing works well with futures system trading combined. Make sure also, you have the means and resources to follow it all day in day out.

11 - Ideally, generate revenues elsewhere for basic living expenses. It is amazing how much this reduces your stress level, and increases your returns... But is difficult to implement.

Turtletrader:

Q1) How much capital do you place on each trade? Is this precise?
Q2) When should you take a loss to avoid larger losses?
Q3) If you begin a losing streak do you trade the same? What formula do you use?
Q4) How should you prepare if trading both long and short positions?
Q5) Is trading affected by commodities that move at different times?
Q6) How is correlation handled in practical trading sense?
Q7) Should you have profit targets? Yes or No?
Q8) Does a portfolio of long and short allow one to trade more positions?
Q9) How is your trading adjusted with accumulated new profits?
Q10) How are stops handled when volatility is a concern?
Q11) Is there a method to limit entry risk with options?
Q12) How does one prepare for unforseen large scale trends?

Money Management questions (Ramon Barros ramon@wr.com.au)

Q1) How much capital do you place on each trade? Is this precise?

Capital per trade is for me a function of my plan's edge, the market's volatility, and the correlation between the markets I trade. I use Gallacher's formula to determine this.

Q2) When should you take a loss to avoid larger losses?

This is a function of my trading plan.
The first element is my initial stop, placed where if hit I am wrong about the trade. Then the second step is, once I have a position, the market must act in accordance with the scenario for my trade - or - as the Phantom of the Pits puts in, the trade must prove that it is a winner.

Q3) If you begin a losing streak do you trade the same? What formula do you use?

Measure a losing streak over the portfolio (not just one market at the time).
I start to reduce numbers once a predetermined level is hit.
Again this is function of statistical data based on my trading history and the current volatility of the mkt.

Q4) How should you prepare if trading both long and short positions?

I'm not sure what this means. Do you mean holding back to back? If so, I don't engage in this practice. For me a long and short of the same position is the same as no position.
If you mean when do I stop and reverse? Again this is a function of my plan.
If you mean some sort of spread strategy - I don't trade spreads.

Q5) Is trading affected by commodities that move at different times?

I trade only FX but keep an eye on Gold, Interest Rates and the Stock Market - using Intermarket Analysis to provide an overview of the longer term trends.

Q6) How is correlation handled in practical trading sense?

This is a function of the capital allocation formula I use.

Q7) Should you have profit targets? Yes or No?

I use the "rule of three" with profit objectives for the first 2 thirds. The last third looks to capture unexpected trends and is usually taken out by a trailing stop.

Q8) Does a portfolio of long and short allow one to trade more positions?

As above - I don't understand "portfolio of long and short..."

Q9) How is your trading adjusted with accumulated new profits?

Once I have made 30%, I add 15% to my capital base and apply the same strategy each 15% thereafter. Each financial year begins afresh.
This means after making 30%, 45%, 60%,... on each step you put 15% of the profit in your account. I want to ensure that I keep some of my profits. Also prevents the normal situation where a loss of 15% is a greater $ amount than a 30% loss because of a rapid escalation of capital.

Q10) How are stops handled when volatility is a concern?

My stops are a function of the market's structure and volatility. If the volatility becomes unacceptable, then I take a trading holiday.
My stops are placed where if hit, I am wrong about the trade. Most times I exit a trade before the stop gets hit. When I plan my trades, I plan for certain things occuring and if they don't, I exit. POP (at http://www.futuresmag.com/library/phantom/phantom.html) puts it much better: "Assume a trade is wrong and has to prove itself within a specific period; if it doesn't exit."

Q11) Is there a method to limit entry risk with options?

I don't use options.

Q12) How does one prepare for unforseen large scale trends?
This is a function of last last third of my contracts.

What is the way to measure the Risk To Reward ratio ??

Really this question depends on a number of factors. I am going to make the following assumptions for this post:

1 you have a written trading plan

2 your plan contains a number of setups

3 you have historical data pertaining to your trades so that you can calculate the mean plus/minus one standard deviation for your profits.

4 you can also calculate the mean plus/minus one standard deviation for your losses.

5 you can calculate the number of trades that have made money and the number that have lost. By dividing wining trades by the total number of trades, you will have the probability of success (PS).

6 you have at least 30 profitable and 30 losing trades from which to make the above calculations.

Once you have the data above, you can now calculate your RR.

The mean minus one standard deviation of your profits is the minimum profit you can expect to make 70% of the time (P); the mean plus one standard deviation is the maximum loss you can expect 70% of the time (L).

The formula for calculating the RR is:

(P*PS)/(L*(1-PS))

e.g.

if your probability of success (PS) = 55%
if your mean -1 std of profit (P) = $1000.00
if your mean + 1 std of loss (L) = $678.00

then your RR =

(1000*.55)/(678*.45) = 550/350 = 1.8

If you have enough trades, you can categorize them by mkts and setups.

You can also substitute the average profit for "mean - 1 std" for (P) and the average loss for "loss + 1 std" for (L). However, I believe that since the average represents a 50% occurrence, you are not putting the odds in your favour by doing this.

If you if need to learn some basic stats and (like me) run at the mention of a maths formula, read "Statistics Without Tears" by Derek Rowntree. A great book in that it teaches basic
stats without a single formula!! I believe all traders would improve their bottom line by understanding basic probability theory.

Note that the formula above says nothing about how many contracts you should take as this is a separate issue altogether. It merely states that given "x" contracts on your past results your RR, 70% of the time, will be "A".

The more volatile the market, the more likely that you will have a large move against you
and therefore you shold work with a tighter stop. However, there also is a contradictory approach that says that the more volatile the market, the more leeway you should give it. What is right and what is wrong? I don't know. You simply have to try both methods and pick the one that you feel the most comfortable with and that matches your overall trading strategy.

The main movement in markets comes from people observing what others around them are doing and reacting to it. So there's a kind of dynamic interaction rather than a specific response identifiable with external news arrival.
It is the dynamic interaction between groups of market participants, who differ in their risk and reward objectives and in their trading time-frames, that is the key to explaining market behaviour.
The fact that some people trade at short time intervals with high risk for profit, while at the other end of the spectrum some trade infrequently at low risk for hedging purposes, creates a set of relative effects where groups react in different ways to events and then react to each other's reactions.

First off, there are many strategies that try and increase/decrease contracts
as a function of the last x trades off of certain equity curve movement.
The point that is important to keep in mind is that any strategy that relies
on the last x trades is inferring that there is a dependency in the trading
system.
For example, if you toss a coin 49 times and all 49 times you get heads,
what do you think will happen on the 50th toss? If the coin is a fair coin,
then regardless of the 49 tosses, the 50th toss still has a 50% chance of
being heads and a 50% chance of being tails. In other words, the results of
the previous 49 trials offers no useful information.
In looking a trading systems, one measure for assessing if you have a
hidden dependency in your system is to do a statistical runs test, also known
as a Z-test. Given the systems Z-score you can determine if you have a
statistical dependence. If you don't (and most systems don't) then trying to
trade based off of the last x-trades offers nothing to improve the system and
perhaps even degrades the performance.
In an effort to try and control drawdown in a money management scheme
I typically pay close attention to the systems historical largest loss on a
single contract basis. If you can incorporate this largest loss in your
money management model, you can usually maintain better control over drawdown.
I prefer to use this number over say average loss, average trade, etc.

Neil Peplinski qa3135@email.mot.com

The best way is to start with a unit bet. Then it all depends on how the hands are streaking. If you are going to win, then you should increase your bet according to the Fibonocci sequence. Otherwise decrease your bet to the square root of the previous bet. This way you will be winning your large bets and losing your small bets. Some call it "locking in" your profits. Everyone knows that everytime you win a hand you are playing with the house's money - so keep betting more, you've got nothing to lose.

Ken Fuchs kenfu@comm.mot.com

go to

Money Management 7

^{back to top}