[AgeKay]

I've found this excellent link in the Is 100% Mechanical Trading Possible? thread that someone shared.

It's a Q&A with an independent retail trader that has automated his trading and trades around 800k to 1.5 million shares a day and make 2cents per trade on average. He made 7 figures profit in 2008 after expenses and taxes.

I used to work as a software engineer and started developing and trading automated strategies in my spare time in 2006. I went full time in 2007 and have been profitable every quarter since

I know it's everyone dream to automated trading, but he gives you insights in how this looks like in the real world. Other than that, he's confirming many things other professional traders say but since retail traders refuse to listen I thought it might be worthwhile summarizing his answers here. (These aren't all his responses just those I found interesting)

So here it goes:

I was making 65k/year for several years. After expenses I was an average middle class employee. I started trading with around 35k and then later borrowed additional money from family and friends.
After taxes, business expenses (I operate through a single member LLC), I earned a little under 7 figures in 2008. I've since ramped up volume, but I'm lagging behind last year.
Yes, I work alone. I am in contact with others traders, but mostly we just bullshit around and speculate on the future of our economy. I'd love to have a few people to bounce ideas off of and to get more perspectives. I've entertained the idea of going past a single member LLC and attempt to expand, but the business is cut-throat and I've only got a few good edges.
I was honestly clueless when I first started trading. I've got to admit I got roped in by the great advertising, flashing lights, indicator loaded charts, and of course the lure of fast/big money. It didn't take me long to abandon all of that garbage. From there on, I, being a computer science guy, decided to stick with automated non-TA based strategies. I was also lucky enough to have an incredibly experienced trader teach me the ropes and a couple of really important lessons. I never looked at the market the same way and developed my strategies upon those principles. It took me around 7 months from inception to have a working trading model with a sizable edge, which I started trading with 26k (down from 35k at start). There were a couple of books that helped me along (On modelling, derivatives, etc), and my stats knowledge also came in handy.

I know of/met a handful of traders who are automated like myself. I'm sure there's quite a few of us out there, but the game is controlled by larger players.

[good internet places where knowledgeable traders can be met] are private and usually by invitation

While I don't post on them, fattail.org, nuclearphynance, and wilmott forums have a good audience. There's also a few gems at elitetrader, but I don't think there's more than a couple of guys who are making money and they don't post often.

Early on, I was in nasty drawdown period and I was having trouble figuring out what was off. I made the same mistake virtually all traders make, I caved to the vast collection of trading psychology books. When the guy mentoring me found out what I was reading, he gave me the following gem: Only pikers worry about psychology, either you have an edge and you exploit it, or you don't have one and you lose and chase every other excuse.

• Trading and Exchanges
• Options, Futures & Other Derivatives
• Option Volatility & Pricing
• Volatility Trading
• Dynamic Hedging

99% of finance books are garbage, but those are the ones I thought helped me in some way or another. There's also plenty of interesting research papers if you've got access to some databases.

I was using a combination of retail feeds, but quickly jumped onto enterprise solutions, because the retail ones just are not sufficient.

I'm using a patched together platform composed of a tick database, message queue, and my own trading framework along with a simple order management system. I'm connected to two brokers via FIX and get data from DTN NxCore.

Any serious broker will have FIX connectivity. The most retail friendly one that still offers FIX, imho, is interactive brokers. Generally, the places that focus on professionals/institutions all do FIX.

I rarely look at bar/candlestick charts. I'll peek at one after hours or when I pull up the S&Ps or something like that. I personally don't put much weight into price patterns/indicators, so I tend to ignore those

I rarely take it beyond simple calculus.

I honestly haven't had good experiences with the prepackaged software. Usually I end up spending more time hacking it to what I want it to do than implementing the strategy itself.

It's an 8-6 job. I'm constantly adjusting, researching, and developing strategies as market conditions quickly change. Anyone sitting on the beach while the trading ATM is printing money wont be in business very long.

I think the quicker you get away from traditional technical analysis, the better. While I think people can make money in many ways, technical analysis is outdated, and we have the computing power and knowledge for better models. I think you need to start approaching the market from a different angle, focus on developing strategies that are not based on charts.

I've used everything from neural nets, SVMs, particle swarm optimization, to artificial immune systems. I have a significant investment in automated news analysis, specifically sentiment interpretation.

I think it should be clear by now that there is manipulation on all fronts. From news, to circle jerk analyst upgrades, to behind closed door bailouts, to front running and flash orders.

I've had almost every aspect of my trading operation give out or fail at some point or another. Redundancy is key.

[On the random walk theory and efficient markets:] I'm in the boat with the behavioral economists.

[On why not every software developer is doing this:] The knowledge barrier has to be overcome. When you want to learn a new programming language, you can buy a book or a course. If you want to become a doctor, you go to medical school. In finance/trading, you can't get access to education, but you still pay tuition. Most don't make it out of the learning stage. You need discipline and motivation, capital, and a lot of time. If it were that easy, everyone would already be doing it.

I'd say about 1/2 of my time I spend on the software/hardware side of things, especially now that I'm trying to improve my infrastructure. I'm something of a jack of all trades, as I have to worry about a lot of different components, so I'm sure most are not optimal. There's no doubt that I could squeeze out more performance and therefore more profit if I hired professional guys with targeted expertise to do specific things. My time could then be focused more on the research/development side, which is what actually makes the money.

Everything is dynamic and the market/participants are always changing. You've got to stay on your toes, there's no room for apathy

Edges are short lived, I'd rather squeeze out every penny I can while I can. I'd sell a model after it quit working, but I'm not a douche bag.

Some of my strategies are fully automated, others are like gray boxes, in that I tune their parameters in real time during the trading day (this makes up about 25% of my trading). I never interfere with any execution, nor do I pull bad trades, as all the logic is handled within the system. Of course on major disaster days like 9/11, I would step in and then close out.
I have an idea of where I want to exit any trade, based on my models, before I put it on; any profit I can get in addition is just a plus.

I use limit orders. I add/remove liquidity on a 50/50 split. I started out trading small size (100 shares), so yes, you can start small.

I personally haven't touched Assembly. I usually try to get away with the least amount of development time as possible, which means I'll start with a high level language like python and then drop to C++ if need be. I've spent significant time optimizing my C++ code. I know that some firms will write their own machine code on specialized hardware.

[On MetaTrader] No. I don't deal with bucket shops or their sketchy platforms

On datafeeds? I pay close to 2 grand for all my data. L1/L2, options and 300 for a futures/fx feed.
For news I pay 450

[On displayed size:]Yes, I use it a lot, although you have to have a good model to decide what is legit and what isn't.

Yes, I utilize a lot of strategies which are generally accepted as bad practices by the retail trading community. I heavily martingale in a futures based strategy.

[On depth in DOM:] Very important, especially in lower volume issues. For example, when I'm trying to push out 100 contracts on the NQ or YM.

People misunderstood how much capital I'm working with, due to my title causing confusion. I do 800-1.5m shares a day, on a 2cent/share average, and made a profit of close to 1m last year. BUT 800k would be plenty for the strategy.
I'm doing this or more size on a martingale type strategy. On the NQ, my maximum loss without hedging is around the 20k mark, which I mitigate with options and can get it down to around a third to half of that. Worst case scenario happens on 1.4% of trades. All of this is automated and I don't tend to stick in issues for 50 ticks at a time; it's a fast game.

I utilize a variety of robust mathematical methods, so an understanding of statistics I would say has been a good part of my success.

There are a lot more factors to consider than just price. Even with just price, think of what you can come up with as far as market neutral strategies goes, possibly combine it with options. I have a non directional mean reversion strategy running on the ES that relies on specific volatility characteristics and I hedge out with options. The strategy works 95% of the time, but the losses are brutal. Therefore the key caveat with this strategy is to contain losses and that's where the options come in.

I started out with stocks because I needed a broad range of issues to trade. I've gotten into SIFs, currency futures, and options since.

Right now, I'm hooked up to IB and Lime Brokerage through FIX. I have used the IB API in the past.

I'm not a market maker nor do I do any equity arbitrage. I'm doing some arbs on vanilla options

[On cause of financial crisis]: People like me are definitely not the cause. High frequency trading is a drop in the capitalistic bucket, a mere fraction of the money traded/made/manipulated. I think you need to go up a few notches.
Historically, the most secure and lodged industries are the financial/banking industries. They'll always be around. There is absolutely nothing wrong with automated trading or short selling. As far as golden parachutes, the value of a CEO is debatable, and I don't think anyone is worth 100million/year.

[StevenSJC]

Great post. Fascinating interview.

I'm not surprised to read that much of what the average retail trader thinks is wrong. Of course, some of this is "one man's opinion" but there are some gems in there.

One for me is the statement about 99% of the books being worthless for him. Or all the various systems. But then he states several books clearly helped him. That it was his introduction to another trader in the beginning that made the all the difference. I think the message here, and from most traders I talk to that no, they did not do it on their own. That yes, they ended up chasing a lot of dubious strategies and system vendors but eventually they hit on one that made a big difference. Even if it wasn't 'out of the box' followed, the culmination of experiences plus hitting on the right book, training, software, etc... helped them turn the corner.

I rarely read about anyone who has done it without any assistance at all. So I think it's trial and error. Once you learn not to chase those selling get rich quick, and instead chase the vendors who seem to be honest and balanced you start to make some real progress. That was how it is for me. To this day I get crazy emails everyday that I ignore, but on occasion I get something that you can clearly see was developed by an actual, real, live trader. Not some bogus marketing or curve-fitted EA. That's what I pursue to expand my knowledge.

Plus boards like TL where you can learn from others as well.

Thanks for sharing.

[SIUYA]

(i wrote this in response to the original thread so added it here as well)

nice find - the few things I found most interesting out of all the discussions was;
- nothing works all the time, there is constant tweaking and research and refinement.
- its traded as a portfolio of systems, with lots of small positions each time
- There is still a lot of work involved on a daily basis, there is no set and forget.
- while they dont look at charts - all price action/movement/charts are derived from the price....so its all the same source - it seems its all a matter of how you wish to visualise it. Either is a chart form - which can cause us to see things that are not really there, or based on actual evidence of things that work as shown via backtesting.
ie; you actually have to be able to visualise the edge yourself and program it, rather than just looking at a chart and saying thats a pattern. (if that makes any sense)
re; emotion/discipline - ideally the testing and the automation should fix all those issues, so long as you thoroughly test, believe and understand the system. AND then be prepared to modify it or drop it, or stick to it when it loses..... all much the same, except like any tool the computers allow some tasks to be made easier.
I personally dont agree with the martingdale strategies, I understand them, but even in his own words, says that the losses are brutal..... I wonder how much impact his hedging, options and spread of strategies limits these losses, while for most people it would ultimately lead to their ruin.
Good luck to them. Its great to see alternative views.

[Kiwi]

An interesting set of views. Good to see an iconoclast interviewed.

I recall being heavily influenced by an interview years back ... and regretting it now to some extent.

One thing he doesn't realize, not having traded thru 911, is that he might not get out of those positions and the hedged black swan will still be interesting. I'd be interested in seeing a followup interview after something like that.

[amswak]

It was briefly mentioned, but I wonder if he ever started that business and is hiring programmers yet...

(p.s. if you are reading this and hiring, pm me...)

[brownsfan019]

Originally Posted by AgeKay »

I know it's everyone dream to automated trading, but he gives you insights in how this looks like in the real world. Other than that, he's confirming many things other professional traders say but since retail traders refuse to listen I thought it might be worthwhile summarizing his answers here. (These aren't all his responses just those I found interesting)

1) It's one person's opinion on trading
2) We have no idea if anything of what he says actually produces live profits
3) There are many, many ways to make money in this business (and plenty of ways to lose as well)

Interesting read? Sure.

Prove that he's a pro and that we all should be quivering at this interview b/c clearly it's superior information? Not quite.

One thing I've learned on my time trading and cruising the forums out there is that there are zero absolutes in trading and what can or can't make money. What you view as no good, others can view as gold.

As for retail traders not listening, I think the new traders and struggling traders do nothing BUT listen to ANYTHING and EVERYTHING and that is their very downfall. Personally I've been doing this long enough that I could care less what some guy on a forum is saying and know better than to treat that as gold or anything new.

In the end we are all taking the exact same information and processing it in some fashion. That's it. How you process that is your decision.

[AgeKay]

Originally Posted by brownsfan019 »

2) We have no idea if anything of what he says actually produces live profits

Have you read his answers / my quotes? Of course, he could be lying but it doesn't seem that he's making that up.

[subterfuge]

Wonder what he is actually looking at and analysing to make his trading decisions?
So very vague, as all 'Non TA' people are.