ETMA_The Evolution

I must admit I really liked the sounds of Basso's approach. It is right down my alley. Using an EMA on ATR makes real sense to me to apply to position sizing. :-)

Fixed Fractional Position Sizing
by Michael R. Bryant

The idea behind fixed fractional position sizing is that you base the number of contracts or shares on the risk of the trade. Fixed fractional position sizing is also known as fixed risk position sizing because it risks the same percentage or fraction of account equity on each trade. For example, you might risk 2% of your account equity on each trade (the 2% rule). Fixed fractional position sizing has been written about extensively by Ralph Vince. See, for example, his book "Portfolio Management Formulas," John Wiley & Sons, New York, 1990.

For information on software for fixed fractional position sizing, click here.
The risk of a trade is defined as the dollar amount that the trade would lose per contract if it were a loss. Commonly, the trade risk is taken as the size of the money management stop applied, if any, to each trade. If your system doesnt use protective (money management) stops, the risk can be taken as the largest historical loss. This was the approach Vince adopted in his book Portfolio Management Formulas.

The equation for the number of contracts in fixed fractional position sizing is as follows:

N = f * Equity/| Trade Risk |

where N is the number of contracts, f is the fixed fraction (a number between 0 and 1), Equity is the current value of account equity (i.e., the value of account equity just prior to the trade for which you're calculating N), and Trade Risk is the risk of the trade per contract for which the number of contracts is being computed. The vertical bars (|) mean that we take the absolute value of the trade risk (risk is usually given as a negative number, so we make it positive).

As an example, consider the series of trades below. The starting account size was $50,000. The profit/loss per contract is shown in the second column ("PL/Contr"). The next column is the trade risk -- $320.28 in this case. The fixed fraction was 0.05 (5%). The next column shows the number of contracts computed according the equation above. Multiplying the number of contracts by the profit/loss per contract results in the position profit/loss ("Pos PL"), which adds to the current equity value to give the new value of account equity, shown in the last column.

http://www.adaptrade.com/images/ffEx1.gif
Trades and number of contracts in an example of fixed fractional position sizing.

Notice how the number of contracts tends to increase over time as the profits accumulate. This can also be seen in the figure below, which shows the equity curve and the number of contracts over a longer span of trades from the same market system. The bar chart, below the equity curve, illustrates how the number of contracts drops after a loss and increases after a winning trade.

http://www.adaptrade.com/images/ffEx2.gif
Equity curve and number of contracts in an example of fixed fractional position sizing.

Fixed risk position sizing can be used to implement another one of Vince's methods, called optimal f position sizing. This is a generalized version of a classic formula called Kelly's formula, which provides the fixed fraction that maximizes the geometric growth rate for a series of trades where all the losses are one size and all the wins are another size. In this case, the optimal fixed fraction is given by the following equation (Kelly's formula, as provided by Vince, Portfolio Management Formulas, John Wiley & Sons, New York, 1990):
f = ((B + 1) * P - 1)/B

where B is the ratio of a winning trade to a losing trade, and P is the percentage of winning trades.

Optimal f position sizing extends the Kelly formula so that the wins and losses can all be different sizes. Optimal f calculates the fixed fraction that maximizes the rate of return for a given series of trades. While this sounds like a good idea, in practice the optimal f value (or the f value from the Kelly formula) often results in drawdowns that are too large for most people to tolerate.

Also, relying on the historical sequence of trades is risky in that the sequence of profits and losses in the future may be less favorable than what was encountered historically. As a result, the drawdowns in the future could be much larger than predicted by the historical sequence of trades. Performing a Monte Carlo analysison the trade sequence is one way to generate a more conservative estimate of the future worst-case drawdown.

A less risky alternative to optimal f is to optimize using Monte Carlo analysis and with a specified limit on the maximum allowable drawdown. This will generally yield a much smaller and therefore less risky fixed fraction than optimal f.

Indeed Bass....indeed.

An old technique adopted in Van Tharp's work also, some think that it stopped working in the current market conditions (the System of random entry with the 3 X ATR Stop not the PS technique) and some say that its still valid, i must confess that i find it more objective than the other PS techniques for it takes the actual asset specific market volatility in account which tend to align us more with the actual market condition be it trending or sideways, while giving us a kind of high probability levels, boundaries to where the asset could reverse or a least stop, which in turn helps us determine a logical and an objective stop loss placement.

I must confess also that (and from a long time ago) i used the ADR and the ATR in my trading and found it very accurate in predicting market reversals, be it Daily, weekly or Monthly and only that tactic had saved me from a lot of stops being hit.

Hi Bass,

Do you still use this system today?

http://www.forexfactory.com/attachme...0&d=1372417192

This is the ETMA which i don't trade anymore, like i mentioned in the first post.

Hi Bass, i found what you wrote .. i asked sine i found this picture posted later than your first post. Anyway, it is clear to me now.

Thank you buddy for ur post, the reason i created this thread is (like i mentioned in the first post) to have a place to park my thoughts and research in writing and to kind of archive those.

I was trading the ETMA with a good edge, since then i evolved a lot and dropped the Short term trading the ETMA (Business oblige, as u may know i have a day job) i kept trading the Medium and long term strategies which i was trading before in the background and which i always post about here, these strategies don't take a lot of my time and i can easily do my job and a bunch of other things in life while trading them at the same time.

Like to mention also that the ETMA system in it self have an edge and its profitable but you have to trim down these indicators a little bit since they are all doing the same job, if i have to trade the system now i would keep the CCI, the ETMA bands and the Donchian channels, as simple as that, i would add one simple rule is to trade only with trend since this is a mean reversion system meaning a both counter and with the trend system, as per my extensive tests the odds are greatly enhanced by doing just that, how to determine the trend remain for u to see, i would use a higher TF or a longer Double MA or a more complicated version of MA which is a Long Term MACD, my test shows that the 75 - 200 is a great combination which have a high expectancy, Mar, Sharpe Ratio, CAGR and the least Drawdown.

Best,
B.R.

CADJPY I'm Finally out.
GBPJPY The tides are turning.
USDCAD Is this the correction coming?, i'm out anyway after the long run.

excellent playz bassramy
nice thread
ill try to slowly grind throu it as your positions!

Nice mate. Get the cigars out and the slippers on. :-)

Thank you 21 and u are welcome.

Hey nerd don't let the steam get to your head, the bitter taste of my small loses is till in my mouth, play defence.
I have a small gift for your nerdy side from your Dear Friend Rob Carver go here Systematic Trading Resources.
His Forum Posts here ( he is wondering what he is gonna do with his already 500K Profits) and his latest Podcast with Michael Covel here, check out his blog also, interesting things there.

It's beginning to look a lot like Christmas......Cheers mate. Thanks for the pressies. That will keep me off the streets. :-)

Risk and R-Multiples

Van says that risk is the amount of money you are WILLING TO LOSE if you are wrong about the market. More specifically, it is how much you’ll lose per unit of your investment (i.e., share of stock or number of futures contracts) if you are wrong about the position you've taken.

This is called the initial Risk, or (R) for short.

One of the key principles for both trading and investing success is to always have an exit point when you enter a position. Trading without a pre-determined exit point is like driving across town and not stopping for red lights—you might get away with it a few times but sooner or later something nasty will happen.

In fact, the exit point you have when you enter into a position is the whole basis for determining your risk, R, and the R-multiples (i.e., risk/reward ratios) of your profits and losses.

Your exit point can be defined in terms of points, dollars or a percentage. For example, William O’Neil says that when you buy a stock, you should get out when it loses 7-8%, while another trader proposes getting out of a stock when it moves 1-2 points against you.

More About Stops

A stop is basically a preplanned exit. Van says that having stops prevents disaster, even though this strongly goes against the grain of the long-term buy-and-hold philosophy.

When the price hits your stop point, you exit the market. A trailing stop basically adjusts that stop when the market moves in your favor, giving you a profit-taking exit as well.

For example, if you buy a stock at $30 and have a 25% stop, you would exit the trade if the price drops 25% to $22.50.

Here's a trailing stop example: Let's say you buy the same stock at $30 (with the initial stop at $22.50), but if the stock moves up to $60, your 25% trailing stop would also move up with it and would be placed at 25% of $60, which is $45.

In other words, you would get out of the trade if the stock turned and dropped to $45, but because you bought it at $30, you would have locked in half your profit, or $15. The trailing stop, in other words, moves the exit point in your favor as the price moves in your favor. However, you must NEVER move it backwards. Thus, if your stock moves down from $60 to $50, you would still keep your exit at $45, 25% away from the high of $60.

In Van’s opinion, this kind of stop is a safe form of buy-and-hold. You could be in a stock for a long time, but if something fundamental changes, it gets you out.

Let's use a real-world example. JDSU went from about $12 in February 1999 to a high of nearly $150 in 2000 (prices are adjusted for a number of share splits). A 25% stop would have kept you in the entire move. You would have been stopped out in April of 2000 at a substantial profit. If, on the other hand, you'd used a buy-and-hold philosophy, you would have suffered severe losses when, in October of 2002, the same stock hit a low of $1.58. To break even, the stock would have to rise 800% from current prices, something it may never do in your lifetime. A stop would have totally allowed you to avoid that fall, along with those of other stocks like Enron and WorldCom before they became headlines.

There are many reasons for using tighter stops, and you will probably need to use them for a variety of different trading styles. We are simply suggesting 25% stops as a substitute for the “buy-and-hold” philosophy.

We are not going to get any further into stops at this point because we want to get back to talking about risk. Just remember, you need to know when you are getting out of a position (your exit point or stop) to determine your risk.

More About Risk, or R

To most people, risk seems to be an indefinable fear-based term that is often equated with the probability of losing; for example, they might think that being involved in futures or options is “risky," or they may be overly optimistic about the trades they make because they don’t understand their worst-case risk or even think about such factors. Instead, they're seduced by trading terms like “options,” “arbitrage” and “naked puts,” or they buy into the academic definitions of risk like volatility, which make for good theoretical articles by academicians, but totally ignore two of the most significant factors of success, the "Golden Rules of Trading":

Never open a position in the market without knowing exactly where you will exit that position,
and
cut your losses short and let your profits run.

Let’s look at the first golden rule in much more detail to be sure that you understand it. It says that you must always have an exit point when you enter a position. The purpose of the exit point is to help you preserve your trading/investing capital. It defines your initial risk (1R) in a trade.

Let’s look at some examples.

Example 1:

You buy a stock at $50 and decide to sell it if it drops to $40. What’s your initial risk?

Your initial risk is $10 per share, so in this case, 1R is equal to $10.

Example 2:

You buy the same stock at $50, but decide that you are wrong about the trade if it drops to $48. At $48, you’ll get out. What’s your initial risk?

Your initial risk is $2 per share, so 1R is equal to $2.

Example 3:

You want to execute a foreign exchange trade, buying the dollar against the euro. Let’s say that $100 is equal to 77 euros. The minimum unit you must invest is $10,000. You are going to sell if your investment drops down by $1000. What’s your risk? What’s 1R?

We made this example sound complex, but it isn’t. If your minimum investment is $10,000 and you’d sell if it dropped $1000 to $9000, then your initial risk is $1000, and 1R is $1000.

Are you beginning to understand? R represents your initial risk per unit. It is simply the initial risk per share of stock, or futures contract, or minimum investment unit. Keep in mind, though, that it isn't your total risk in the position, because you might have multiple units.

What’s My Total Risk?

Your total risk would be based on your position sizing and how many shares or contracts you actually buy.

Let's say you bought 100 of the shares in Example 1, which would be 100 multiplied by the share cost of $50 each, giving you a total COST of $5000. However, you are only willing to risk $10 per share, so $10 multiplied by 100 shares = $1000 total risk for this position.

If you bought 100 of the shares in Example 2 at $50 each, your total COST would be $5000, but because you plan on getting out if the stock drops to $48, your total risk is $2 per share multiplied by 100 shares. In other words, you'll only be risking $200 of your $5000 investment.

Understanding R-Multiples

The next key point for you to understand is that all of your profits and losses should be related to your initial risk. You want your losses to be 1R or less—which means that if you say you’ll get out of a stock when it drops from $50 to $40, you actually GET OUT when it drops to $40. If you get out when it drops to $30, your loss is much bigger than 1R. It’s twice what you were planning to lose, or 2R. You want to avoid that possibility at all costs.

Instead, you want your profits to be much bigger than 1R. For example, you could buy a stock at $8 and plan to get out if it drops to $6, so that your initial 1R loss is $2 per share. If you go on to make a profit of $20 per share, you've made 10 times what you were planning to risk—in other words, you've made a 10R profit.

You try it:

1. You buy a stock at $40 with a planned exit at $35. You sell it at $50. What’s your profit as an R-multiple?

2. You buy a stock at $60 and plan to get out if it drops to $55. However, when it goes that low, you don’t sell. Instead, you just stop looking at it and hope it will go
back up. It doesn’t. It becomes part of the headline business news involving corporate scandal and eventually the stock becomes worthless. What’s your loss as an R-multiple?

3. You buy a stock at $50 and plan to sell it if it drops to $49. However, the stock takes off and jumps $20 in three weeks when you sell it. What is your profit as an R-multiple?

Answers

1. A 1R loss is $5. Your profit per share is $10, so you have a 2R profit.

2. A 1R loss is $5. Your loss per share is $60, so you have a 12R loss. Hopefully, you can understand why you never want to let this happen.

3. A 1R loss is $1. You profit per share is $20, so you have a 20R profit. And hopefully, you understand why you want this to happen all the time.

What's really interesting is that once you understand risk and portfolio management, you can design a trading system with almost any level of performance. For example, you can design a system to trade for clients that would make about 30% per year with only 10% drawdowns.

On the other hand, if you want to trade your own account and be a little more risky, you can design a system that will produce a triple-digit rate of return as long as you have enough money to do so and are willing to tolerate tremendous drawdowns.

It’s a whole new way of thinking for some, but most successful traders think in terms of risk/reward, which, of course, gives them an edge out there in the markets. Learning to trade and invest in this way will keep you in the game longer and enable you to run with your profits and cut your losses short. What could be better than that?

lets talk a bit concerning win rate and RR etc.
ya, we all know that with 1:4 RR splendid win rate is at 35% easily. Same goes for the 1:1 RR with 65% win rate .

HFTs carry very very high probability of win rate 95++%? whereas they hold a negative RR ratio. But we aint talking about that neither.

Im just talking that the notion of getting stuff right (as in win rate) does not neccessary have negative effects as said in this article. When I started more in mechanical and algorythmic ways of trading this was my starting point definately. When started I was aware that I need to start somewhere. So decided to start at the basic common sense principle: win rate on 1:1 RR.
When I established (still establishing/new generations etc) a system from up to bottom my agenda was to definately get the best entries possible on the 1:1 RR and from then on I could move forward in testing and strechin/skewing PL/RR in my direction.. (skewing RR in my favor). But one needs to start somewhere eh.

With mechanical system this is easily done than in prop trading. But im not really into prop/manual trading anymore. And anyone asking me how one should handle trading i tell then to go directly into mechanical stuff and algorythmic trading. Too many advantages lies in this style of trading.

Well, to get back to higher RR ratios in traders favor. Well, Im finding this as a new chapter in the testing environment and in the forward testing statistics. But what Im seriously thinking about is that while 2nd generation of pigz (my EAs are called pigs) is trying to really catch the best trades plus its very much structured as a statistical trading equipment
- i figured that one needs/tries to program the algorythmic principle of monitoring the open trades. (hard stuff to do and Im not talking about complex trailing stops etc). Im talking about monitoring the moment of now every minute or so to really not leaving the trades unattended.
With pretty much fixed SL/TP at desired ratio in the system, ones advantage to have dynamic ratio in the system should be immense.
That is what im thnking about lately and trying to tackle this issue in mechanical way again.

Appreciate any view.

Position Sizing
What is it?

Money management is a very confusing term. When we looked it up on the Internet, the only people who used it the way Van uses it were professional gamblers. Money management as defined by other people seems to mean controlling your personal spending, or giving your money to others for them to manage, or risk control, or making the maximum gain—the list goes on and on.

To avoid confusion, Van elected to call money management "position sizing." Position sizing strategies answer the question, "how big should I make my position for any one trade?"

Position sizing is the part of your trading system that tells you “how much.”

Once a trader has established the discipline to keep their stop loss on every trade, without question the most important area of trading is position sizing. Most people in mainstream Wall Street totally ignore this concept, but Van believes that position sizing and psychology count for more than 90% of total performance (or 100% if every aspect of trading is deemed to be psychological).

Position sizing is the part of your trading system that tells you how many shares or contracts to take per trade. Poor position sizing is the reason behind almost every instance of account blowouts. Preservation of capital is the most important concept for those who want to stay in the trading game for the long haul.

Why is Position Sizing so Important?

Imagine that you had $100,000 to trade. Many traders (or investors, or gamblers) would just jump right in and decide to invest a substantial amount of this equity ($25,000 maybe?) on one particular stock because they were told about it by a friend, or because it sounded like a great buy. Perhaps they decide to buy 10,000 shares of a single stock because the price is only $4.00 a share (or $40,000).

They have no pre-planned exit or idea about when they are going to get out of the trade if it happens to go against them and they are subsequently risking a LOT of their initial $100,000 unnecessarily.

To prove this point, we’ve done many simulated games in which everyone gets the same trades. At the end of the simulation, 100 different people will have 100 different final equities, with the exception of those who go bankrupt. And after 50 trades, we’ve seen final equities that range from bankrupt to $13 million—yet everyone started with $100,000, and they all got the same trades.

Position sizing and individual psychology were the only two factors involved—which shows just how important position sizing really is.

How Does it Work?

Suppose you have a portfolio of $100,000 and you decide to only risk 1% on a trading idea that you have. You are risking $1,000.

This is the amount RISKED on the trading idea (trade) and should not be confused with the amount that you actually INVESTED in the trading idea (trade).

So that’s your limit. You decide to RISK only $1,000 on any given idea (trade). You can risk more as your portfolio gets bigger, but you only risk 1% of your total portfolio on any one idea.

Now suppose you decide to buy a stock that was priced at $23.00 per share and you place a protective stop at 25% away, which means that if the price drops to $17.25, you are out of the trade. Your risk per share in dollar terms is $5.75. Since your risk is $5.75, you divide this value into your 1% allocation ($1,000) and find that you are able to purchase 173 shares, rounded down to the nearest share.

Work it out for yourself so you understand that if you get stopped out of this stock (i.e., the stock drops 25%), you will only lose $1,000, or 1% of your portfolio. No one likes to lose, but if you didn't have the stop and the stock dropped to $10.00 per share, your capital would begin to vanish quickly.

Another thing to notice is that you will be purchasing about $4,000 worth of stock. Again, work it out for yourself. Multiply 173 shares by the purchase price of $23.00 per share and you’ll get $3,979. Add commissions and that number ends up being about $4,000.

Thus, you are purchasing $4,000 worth of stock, but you are only risking $1,000, or 1% of your portfolio.

And since you are using 4% of your portfolio to buy the stock ($4,000), you can buy a total of 25 stocks without using any borrowing power or margin, as the stockbrokers call it.

This may not sound as “sexy” as putting a substantial amount of money in one stock that “takes off,” but that strategy is a recipe for disaster and rarely happens. You should leave it on the gambling tables in Las Vegas where it belongs.

Protecting your initial capital by employing effective position sizing strategies is vital if you want to trade and stay in the markets over the long term.

Van believes that people who understand position sizing and have a reasonably good system can usually meet their objectives through developing the right position sizing strategy.

Position Sizing—How Much is Enough?

Start small. So many traders who trade a new strategy start by immediately risking the full amount. The most frequent reason given is that they don’t want to “miss out” on that big trade or long winning streak that could be just around the corner. The problem is that most traders have a much greater chance of losing than they do of winning while they learn the intricacies of trading the new strategy. It's best to start small (very small) and minimize the “tuition paid” to learn the new strategy. Don’t worry about transactions costs (such as commissions), just worry about learning to trade the strategy and follow the process. Once you’ve proven that you can consistently and profitably trade the strategy over a meaningful period of time (months, not days), you can begin to ramp up your position sizing strategies.

Manage losing streaks. Make sure that your position sizing algorithm helps you reduce the position size when your account equity is dropping. You need to have objective and systematic ways of avoiding the “gambler’s fallacy.” The gambler’s fallacy can be paraphrased like this: after a losing streak, the next bet has a better chance of being a winner. If that's your belief, you'll be tempted to increase your position size when you shouldn’t.

Don’t meet time-based profit goals by increasing your position size. All too often, traders approach the end of the month or the end of the quarter and say, “I promised myself that I would make “X” dollars by the end of this period. The only way I can make my goal is to double (or triple, or worse) my position size. This thought process has led to many huge losses. Stick to your position sizing plan!

We hope this information will help guide you toward a mindset that values capital preservation.
I've talked to many folks who have blown up their accounts. I don’t think I've heard one person say that he or she took small loss after small loss until the account went down to zero. Without fail, the story of the blown-up account involved inappropriately large position sizes or huge price moves, and sometimes a combination of the two.—D.R.Barton, Jr.

ok, this is my goal really.
but, im really lousy programer. 0,5 / 10.
so initially have a big problem. im still in the process to collect a group of profiles which will work also on this issue.
atm all my testings are forward testing. Some 2y running now.

seems you test on different environments means you have relatively simple system rite. so you can translate it into different languages.
im still in MT4 environment where backtesting is really a joke. Especially as im on the currency index terriotory and when trading XXX/YYYY pair i look at the whole market and not at particular pair only.

Any solution you saw in your testing with Multicurrency index trading and multy TF stuff

thanks later

21., you don't have to really be a programmer to build a robust trading system these days, all the simulation and backtesting platforms (best software for backtesting futures strategies?) have a drag and drop kind of mechanisms, even the most open structure ones like right edge, now i'm not a programmer too but i am a Computer and software expert (it Means for example that i memorized Windows 95 to the last Win 10 and i'm very comfortable with Hardware) and this is a prerequisite to Build systems.

As you can see from these links above, i too decided at last on two Backtesters, the Most expensive Trading Blox for a stand alone systems and Ninjatrader because i use it for trading too but its a little bit slow, if you wanna stay on the simple side i heard a lot of good stuff on Amibroker and WealthLab, Tradestation is OK but i never tried it and Metatrader is not an option.

Like i mentioned Before i Backtested, optimized, walk forward tested 63% In sample and 33% out of sample Data to see if the system behave like it should and then Tested the robustness of the system with a Monte carlo Analysis, (Monte Carlo simulation refers to process of using randomized simulated trade sequences to evaluate statistical properties of a trading system thus allow to verify robustness of the trading system, find out probability of ruin and many other statistical properties).

I then apply an incubation period of 6 - 12 Month to make sure the system works well.

​My portfolio consist of 38 - 42 Instruments in different asset classes, each Instrument has passed this entire work process and the portfolio has been backtested entirely to verify the end Results, all backtesting been done with Fixed Fractional position sizing of 1% taking in consideration the Volatility of the Instrument to Normalize the Dollar Value of the Portfolio or the Dollar risked in every trade, so High volatility Instrument tend to trade less contract and vice versa.

Taking in consideration the Market conditions i have found Currencies (Except some majors) and Bonds to have higher ATR's which means more volatility than the Stock Indices thus they tend to Whipsaw more (Indices and because of Fundamentals being doing this lately and that's why last year was so difficult to trade and trend followers did poorly) Janet Yellen gave the CTA's a hard time last year because of the speculation of the Monetary policy tightening.

​This is hard work i confess but in my opinion this the path if you want to be in the 5 - 10% of profitable Trading Businesses and unfortunately there is no shortcuts, u have to analyse and evaluate your opportunity cost and see if its worth wasting your time or not, IMO i won't waste my time doing all this process (and we didn't talk about the other aspects of testing yet, Data etc...) if i don't have at least 100K to trade.

Hope it helps, Best.
B.R.

A sample of my last Double Donchian Optimisation process.