entry/exit strategies: two schools of thought

School 1: Cut losing trades short, let winning trades run

1. Set close stop-loss points, and distant take-profit points (or take profit only upon pre-set exit signals)
2. Either a) high percentage of failed trades or b) very few trades (ie clearly defined entry signals)
3. Less emotion involved, since there won't be any large running losses to worry about. But many failed trades may raise questions about chosen strategy

School 2: Take profits consistently

1. Set close take-profit points, and distant (or no) stop-loss points
2. Allows dollar-cost averaging (enter more trades when the price goes against you, in anticipation of a reversal. In the most extreme case, a Martingale strategy can be applied, where the size of each subsequent trade is enough to cover all preceding losses in event of a reversal)
3. Account statement will always look good, since losses are rarely realized on paper (ideally)
4. Emotionally, running balance may be cause for concern. Since profitable positions are closed quickly, open positions will more often than not be in the red.
5. Eventually, a Black Swan event will come along, which may ruin the entire account

Any other schools?

ECN vs trading desk

Received an email by Alex Nekritin from traderschoicefx.com. I'm presently biased against dealing desks (I hate stop hunters!), but this article provided me with a different perspective. Read on:

Many people are concerned about going to an ECN broker and not trading through a deal desk. In this article I hope to shed some light about how this works and what to look for when selecting a broker to make sure that you don't become a victim of un-just dealing practices.

An ECN dealing model allows the many market participants to execute trades with each other through an electronic network. That's what an ECN stands for electronic communications network. As you know forex is a zero sum game so for every winner there is a loser, for everyone going long there is someone going short. So what an ECN does is match up your order with the order of another market participant. You are probably asking yourself the same question as I asked myself when I first found out about an ECN. Will there be a seller every time I am buying? Well there are market makers and banks in the ECN that are consistently taking on trades and laying of the risk. They may have their own buy and sell programs that they are trading on. These banks allow clients to get better liquidity and tighter pricing in the ECN. Some benefits of using an ECN is you get anonymity as the other participants do not see who is trading on the other end and cannot flag your account and trade directly against you. Another benefit is you can make your own market, you can place orders in between the bid and the ask.

Some forex dealing firms use a dealing desk approach. With this approach their desk acts as a sole market maker and takes all long and short positions on. The desk has certain risk parameters that have been set up. And based on these calculations the aggregate net position of the dealing desk is hedged of. So if the desk itself is net long or short a certain amount of EUR/USD for example they will take a trade of that amount in the opposite direction with a liquidity provider. If all the clients are net long 1 billion EUR/USD, the desk will go long 1 billion EUR/USD and thus have a hedged position. So for every pip they lose in aggregate to their clients they will win on their hedge. Its obviously not as simple as I just explained it but that's the basic nature of the dealing model. Some of the advantages of going through a deal desk are that you always know your transaction costs as the spreads stay fixed, you know who the counterparty will be every time in case you need to get issues resolved. There are however some disadvantages as well, the dealer will always know who you are, and you cannot go in between the bid and the ask.

Although many people are strong proponents of the ECN model which does seem a lot more transparent. The dealing desk approach can work as well, as long as you are trading at a well capitalized firm with numerous deep liquidity relationships. The bottom line is you want to make sure that everything about your dealing firm, platform and overall trading set up is a fit to your needs. However there are some things you need to check right away to make sure that you are trading at a solid firm. Because what good are tight spreads if you can not withdraw your money at the end of the year. :) Take a look at our site for example to learn more about forex trading brokers.

In general first find out the firms capitalization. You can find this at http://www.cftc.gov/marketreports/financialdataforfcms/index.htm. You want to make sure that the firm is well capitalized, I would say over $25 million for adjusted net cap is a good start. This means that they have enough money to have solid liquidity relationships. Next I would actually ask the firm who their liquidity providers are. You want to make sure they are big firms like JP Morgan or Bank of America. Some firms may claim they have no dealing desk on their websites, but in actuality send all their order flow to another dealing desk, you need to be really careful about that. In this case you would be much better of using an IB as you can lower your transaction costs by receiving a volume rebate and trading at the source. One way to check this quickly is an ECN will always have floating spreads. Spreads cannot be fixed at an ECN. So if somebody is offering fixed spreads and saying they have no dealing desk. Guess what? They are going to another dealing desk.

Regression to the Mean

Here is a statistical principle that, to me, works almost like magic when I looked at it the first time:

Regression to the Mean

According to an article on Wikipedia, if I know the mean or expected value of a random variable (coin flips, test results, etc), if I take a random result from the variable's sample space and it turns out to be far from the mean, (either beyond the 25-percentile on the distribution curve, or at a certain distance beyond the standard deviation), if I take another independent sample, the second value will be much closer to the mean, depending on how independent my second measurement is from the first.

Does regression to the mean occur in stock price movement? Or perhaps this is better applied to forex spot prices, due to its greater volatility which more closely approximates normal distributions (will have to find out whether this is true). If that's the case, one should simply go long when the price is far below the mean, and short when its above.

The key reason why this doesn't apply is because the range of values of the second price is almost always dependent on the first, since the price-time graph is continuous. But instead of being applied to the instrument price as a whole, we can use it to gauge the likelihood of more specific events which are more random, eg after the MACD crosses over and this indicator goes up and so-and-so, what is the average price behavior over the next 10 candlesticks? In this case we might simply open a position based on the average behavior, But if regression to the mean applies, we can also wait for a large deviation to occur in the negative direction, which according to the principle will increase the odds that this won't happen the next time.

This brings me to the question: what is the best way to model the price of a financial instrument as a random variable, and hence in a probability distribution graph?

First Post

I'm not a gambler. I won't get excited at low-probability lotteries. I dislike betting on football, as it takes the fun out of supporting my favorite team (if someone offers me good odds for Spurs to lose against Arsenal, will I take it?). I like easy money as much as the next person, but I don't believe in casinos. All casino games are designed such that the dealer has the advantage, so no one can ever win in the long run, so the most you'll get is a short-term thrill from beating the odds, in which case you might as well go to the lotteries.

But financial markets are a different thing altogether. It's a matter of company and/or economic fundamentals, technical signals, and ultimately, how your fellow traders feel about the same stock. Are there millions of them, or is it only a few thousand traders whose large transactions matter? In any case, they are the cause of all price movements, and consequently the changes on your profit/loss statement. Do the laws of probability dictate that the odds are firmly set against day traders? Is it possible to retrieve sufficient parameters to calculate the probability of a making successful long trade?

And so, thinking that I know all there is to know about forex after reading a few Investopedia articles, I've put $3000 of my savings into a forex account. And so I've lost $3000 within a week. And I've allowed a unit trust to lose 30% of its value due to the recent downturn. That sums up my experience with the financial markets.

But I'm not giving up. You have charts, and all kinds of technical indicators out there which can allow you to predict future movements. The only thing is, there's no fixed rule saying how they should be used. An MACD crossover indicates a change in trend. Does that mean you can open your long position now? Or shall I wait for the Stochastic Oscillator to crossover in order to confirm the trend? How about RSI? Has it crossed 50, or is it already in the 70-80 range, which means you're probably too late to enter a long position. Or does it mean that there's a strong trend and that you should hop in? What about the ADX? What about the Aroon oscillator? Wait, what was I talking about?

The key questions to be asked, I feel, is:

1. Given the economic/technical/whatever signals, what is the probability that the price will move in a given direction, and that I will profit from this movement?

2. How will this probability be affected with different take-profit and stop-loss parameters?

Once a situation is arrived where the probability is known for certain, or perhaps a range of probabilities, or perhaps a situation where the probability is above 50%, we can conclude that consistently entering a trade whenever the given situation occurs will result in profits in the long run.

It'll be cool if this can be done mathematically. I've taken a couple of classes in probability and statistics, and was enthralled by how the theorems and distributions can be used to explain and even predict real-world events. Hopefully, perhaps in combination with technical analysis, perhaps it will be possible to predict future chart prices using these methodologies. Perhaps.