About the experiment

Over 5 years ago I graduated from the University of Florida with an MS and a BS in Nuclear Engineering. I then moved to the Pittsburgh area to work at a large multinational engineering company. The first few years of my career, I worked as a technical person cranking out work directly related with my field of study. It didn’t take long, however, to realize that the best way to advance my career would be to pursue a path in Project Management, so I left my technical work to become a Project Engineer on a path to becoming a Project Manager.

Project management is great career-wise, but it essentially renders all my earlier knowledge useless. I have to learn a new set of skills, mostly soft skills that come naturally to most people other than engineers. I’m still convinced that this was the right decision, however, not just for me but also for my family’s well-being.

That is what brings me to this blog. In an effort to secure a better future, I’ve contributed diligently to my 401k. That same 401k became a black hole in 2008-2009. Like most people during this financial crisis, my trust in the stock market as a viable investment vehicle has been completely shattered. It could be going along just great, meeting your 10% yearly return projections, and then BAM! — the rug gets pulled out from under you and you just lost 50% of your nest egg.

So where am i supposed to grow my investment? The stock market is rigged, bonds give you next to nothing, CDs are a joke, munis could default any minute, and if I just hoard cash, it’s only a matter of time before the government cranks up on hyperinflation to debase the debt and render my savings worthless.

That’s where, in my long search for a viable investment vehicle, I came across the Forex market. Forex is the spot market for currencies. You essentially trade currency in pairs, betting that one will rise (or fall) against the other. I’m not going to go into detail about this, since there are some excellent sites out there that tell you all about the Forex market, but I am going to discuss my explored approach to Forex trading as a retirement investment vehicle.

Now if you know anything about Forex, you’ll know that the words “safe investment” do not belong anywhere near a sentence with the word “Forex” in it. This is because Forex trading can be extremely risky, and you can lose all of your money in a matter of seconds if you don’t know what you’re doing. In a currency pair, the price changes in increments of $0.0001, called a “pip.” Depending on your position size, a pip could be worth anywhere from one cent to $10, $100, $1000, etc. The only limiting factor is the size of your deposit and the leverage at which your broker will allow you to play. The higher your position size, the closer your bet gets to your maximum leverage, and the more likely it is you’ll get a margin call if the price goes against you. Again, there are excellent sites out there that go into detail about this, but the point I’m trying to make is that you can easily dial the risk to insane levels in Forex. On the other hand, for someone who knows what he’s doing, the variable position sizing can become a great risk management tool.

So if Forex is actually such a terrible investment (or not even an investment at all) and so inherently risky, why do I keep going on and on about it? As it turns out, most of the trading in Forex is done by robots. A Forex robot, also called an “Expert Advisor” or EA for short, is nothing more than a computer program that places trades when certain conditions are met. These conditions are typically based on states of different technical indicators (e.g.: the slow EMA crosses the fast EMA, the RSI below 30, etc). There are tons of people online selling these robots, and as you can guess, the vast majority of these are scams. You can easily develop a robot based on some simple trading rules, optimize it on some historic data by tweaking its parameters (e.g.: EMA period, RSI period, etc) and it will give great backtested results over that specific past history. Another common trick is to build a robot that will enter the market at random times with a TP of 1 or 2 pips and an SL of 500 pips. These will typically show huge profits in an idealized, zero-spread, instant execution environnment (which is fantasy). Then the scammers put up the backtest data as proof that their amazing robot will give you 1000% daily returns or some such nonsense, you shell out hundreds of dollars, and watch your account disappear when you actually use the EA on a live or demo (preferably) account.

So what should I do? Trust some Russian webpage promising me an infallible EA? I have better sense than that. Remember all that knowledge I have that is now useless in Project Management? Well, it may actually come in handy after all. In college I learned how to write Monte Carlo code to sample the path of subatomic particles in scattering and absorbing media, and how to code this from scratch in C++. In addition, I used Monte Carlo methods extensively to design nuclear systems and radiation shielding. I have the programming knowledge and scientific methodology to design and implement a trading system based on Monte Carlo sampling, so why not give it a shot?

In my research I was unable to find any commercially-available EAs that use the Monte Carlo method, and I took this as a very good sign. If the scammers aren’t hawking this, it must mean that a Monte Carlo-based trading method might actually work.

Before i get too excited, though, it’s important to recognize some limitations of this approach:

  • The EA will be collecting historic data from the trading chart to assemble the statistics that define the sampling functions powering the EA’s predictive abilities. This assumes that the market will continue to behave the same as it has in the past, and we know this is not always the case.
  • The nature of Monte Carlo methods requires a large number of samples, so continuous monitoring is not possible. The time scale being used cannot be smaller than the amount of time it takes to calculate a projection. Monte Carlo is very computationally expensive, and in this case it requires a lot of historic data to assemble meaningful statistics. A somewhat fast computer and good source of historic data will be required
  • News releases and other events can cause large swings in the Forex market. My EA will be particularly exposed since it has no defense against this. In addition, a change in central bank interest rates, or fiscal policy can alter the operation of the Forex market, invalidating historical data. An emphasis must be placed in recent data, and if some such event takes place, time should be allowed for the statistics to redefine themselves.

After recognizing these limitations, I still think I may have a shot. Of course, I’m not putting any real money in at this point. We’ll start our grand numerical experiment with a $500,000 demo account and see what happens. Will it soar into the stratosphere with my nuclear powered EA, or will it have a catastrophic meltdown?

Only time will tell.

Advertisements

3 Responses to About the experiment

  1. Firoz says:

    Well thought out. Wish I could use my Six sigma black belt knowledge to compile some stats and code. Was thinking of some sort of regression/predictor -corrector method.

  2. gatornuke says:

    I should mention that my approach has changed somewhat since I wrote this. As should be evident in other parts of my blog, I’ve embarked on a series of experiments to better refine my EAs, and I plan to ultimately use them in a vehicle that allows others to participate.

  3. Pingback: Anonymous

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: