sappjason

I picked Maryland, because I live in Maryland. I actually live about 10 miles from my co-location facility. I don't have any need for sub-second execution. My main need was to have a server that is up and running every day without interruption, versus running it from my basement. If you have the same type of needs, I would recommend just choosing a local co-location facility and go with that. Jason

I consider myself a successful trader, but I am not full-time. I am a software engineer, but have an automated trading system that I busted my ass on for about 2 years. Been running since 2008 (started with $30K) and slowly added more and more money as I proved to myself (to the best of my ability--via backtests and forward tests) that what I'm doing actually works. MightyMouse said the following: "I also think that it isn't as lucrative as people think". I think I'm starting to slowly agree with that statement as I've been doing this successfully now for more than 3 years. The jury is still out on this, so time will tell. Now, just a little background. I tried to trade (as a hobby) for 10-15 years and pretty much just pissed money away year after year after year. I blew up badly once when I lost $27000 in about 1 hour. I had a 2nd blow-up, but it wasn't as bad. After many years, I finally had a "trading breakdown" when I went through some serious emotional anguish over the fact that I was a "trading failure". I finally faced myself that I couldn't trade my ass for a hole in the ground and decided to buckle down and start to use my software development skills to mine data and look for some semblance of profitability in the markets. I did this for many, many months and thought I had finally discovered something that may (just may) work. Once I started acting on this strategy with real money, I found out over the course of a couple months that I was wrong. It was through this (much more disciplined) approach that I stumbled onto what I'm doing now. But this next step took more than a year of concerted effort (on my part) where I sacrificed one hell of a lot of family time and busted my ass constantly until I finally "cracked the nut". I still had a long way to go, but I finally had the underpinnings of something that seemed to work. I worked another solid 1 to 2 years honing and tweaking my strategy to what it's become today. Even with all of this blood, sweat and tears, I only have a winning strategy that has an EV of .10 and this is all I've got!!!!!! One automated strategy that supplements my income. My desire (of course) is that this will eventually replace my income and I finally think I'll truly get there, but I'm not counting my chickens until my trading consistently brings in north of $200K per year (for at least two years). Jason Here are two charts of my actual profitability since the inception of my automated trading back in 2008. Not a ton of money, but I'm pretty damned proud of it!!!!! Daily Results http://postimage.org/image/2xst806mc/ Trade by Trade Stats http://postimage.org/image/2xs2rempw/

Yep, I agree with MightyMouse as well. I trade a bit differently in that my computer either takes me out or does not. My problem lies in the fact that I friggin' watch my system like a hawk and cause myself unneeded stress, because there's nothing that I would do to any of my positions either way..... Good luck..... Jason

I'm not a discretionary trader, but rather an automated day trader that trades exclusively US equities. So, as far as your position sizing questions are concerned, I'm just going to state the obvious. Here it goes: If you determine that you want to go long stock XYZ at $50.25, you must also determine exactly where you are willing to get out for a loss. Lets say that you determine your stop loss location to be $49.75. This gives you a $.50 cent stop size. If your max risk is $100, then you should purchase 200 shares at $50.25 and place your stop at $49.75..... Lastly, your exit location (for your profit) would be $51.00 (for a profit of $150). Now, with all of that said, there are other costs involved here that you already know about. Namely commissions and slippage. In most cases, commissions are easy to determine, but slippage is the hard part (especially on your stop). As your position sizes grow, slippage will eat away at your bottom line much more than commissions (I'm at this point with what I'm doing now). In fact, with any momentum based system that needs to rely on market orders for entry (where position entry is more important than price), you will eventually hit a brick wall where your percentage returns will diminish to the point where trading that way no longer makes sense. Back to the matter at hand..... If you purchase your 200 shares at $50.25 and it hits your stop location of $49.75, your real risk is actually more than $100. It will be $100 + commissions + slippage (on your exit and entry---if you used a market order to get into your position). My ATS software estimates how much it thinks it will slip the position it's about to enter and then size the position down to account for slippage on entry and exit (I use market orders for entry and mostly stops for exit, so this incurs a lot of slippage). My risk size is currently about $600 per trade and because I'm a day trader, my estimated slippage is typically kind-a high. Therefore, many times, my system sizes the position size down quite a bit, so that my commissions and estimated slippage never account for more than 10% of the total risk size. In other words, if my desired max risk size is $600, I don't want slippage and commissions to add up to more than $60. Lastly, moving your stop after your position is in the green by a certain amount is not a bad thing, but it would really be helpful to you if you understand how this stop movement affects your EV over the long haul. In other words, have you back-tested what you are doing to determine if you think your system is actually profitable (over the long haul)? Jason

Slight correction... The T&S shows trades that have occurred and at what price. A trade does not always have to occur at the "visible" bid/ask. You could have $22.10 bid x $22.15 ask and have a trade tick off on the T&S at $22.13. A complete T&S will also show all bid/ask price and size changes as well. Jason

I lease rack space at a co-location facility so that I have zero power outages and 100% internet up-time. For me, it has nothing to do with low latency as I'm not doing any kind of HFT. During the past year, I had one interruption and it was during an evening for about 5 to 10 seconds. When I first started this ATS, I ran it at my house, but you would not believe the silly stuff that happens at your house when you really want to achieve zero interruptions from 9:30 to 4:00. Here are a few of the more laughable things that happened to me when I ran my ATS at my house: - I had a hamster chew my phone line (I used DSL). - Verizon switched my phone number (without my consent). - A bulldozer hit a power line while doing construction in the field behind my house. - My son unplugged my power strip to plug in something of his own. The moment of clarity that lead me to finally move my PC out of my house to a facility was when I missed out on being able to close my positions at the appropriate time. This small problem (that happened at around 2:00ish in the afternoon) caused me to miss out on $4300 of unrealized gains. Needless to say, this simple mistake would have paid for 3+ years at my co-location facility. So, eventually, the math was a no-brainer..... At this point, the amount of money this has saved me can now probably pay for the next 30 years of my co-location fees.... Jason

I run my own fully automated trading system. It's been running since early 2008. I have a 1U rack mount server with dual processor 3 GHZ (dual core), 4 GB ram and two small 200GB hard drives (with external drive for backup), running 64bit Windows Server 2008. I run my system in a small co-location facility in Maryland and I pay $150 per month for my co-location service. I also pay $75 per month for IQFeed real time data and $20 bucks per month for a FREE End of Day Stock Quote Data and Historical Stock Prices subscription for eod data. I use Interactive Brokers and developed my entire system in Java against IB's TWS API. My system trades only US Equities and does around 13 round turns per day. I blow through around $35K in commissions per year. With all of that said, the last thing to worry about is your PC strength, hard drive size and video card.... Your #1 priority should be developing a positive expectancy, consistently profitable trading system. This step (and nothing else) should be your sole focus for AS MANY YEARS as it takes.... Trust me, this step alone, many never achieve.... Jason

People set the bid/ask prices. That's the simple answer. Sometimes its you and me simply using limit orders. Other prices are set by the various market makers... Many times these days, its computers that are programmed by people, but the bottom line is that its people using LIMIT ORDERS... You can see this yourself if you want. Pick any stock. Bring up level 2 quotes and enter a limit order way outside the market. If a particular stock is trading at $10.00 / $10.01, then for your test, enter a limit buy order for 100 shares at $8.00. You will see your order appear in the book.

The guy at $10.04 with 2200 units would get his fill when enough buyers enter the market (or current buyers shift their prices upwards for whatever reason) until the $10.04 ASK gets exposed and people start eating away at his 2200 units with buy orders (market or limit buy orders). When you say a "market maker" lifts the bid/ask prices, I'm assuming you mean that the MM just moves his current bid/ask orders (or adjusts their prices). But, its an entirely different story if the MM has $10.00-500 BID and $10.01-1000 ASK and someone comes into the market with a 5000 BUY MKT order. In this case, the $10.01 x 1000 shares would be lifted, because they were BOUGHT by the 5000 MKT ORDER. Therefore, the $10.01 ASK price would disappear (exposing the next best ASK price above $10.01 which doesn't necessarily have to be $10.02---it could be $10.06----whatever is in the book at the time), because the 5000 share MKT ORDER executed against them.

I'm going to attempt to explain the actual mechanics of how an instrument increases/decreases in price, so bear with me, this may get confusing. Also, I'm not going to consider hidden orders in this example at all, but they most certainly exist (at least in equities markets). So, for this simple example, we have a fictitious stock with a symbol of ABC and the current order book looks like this: A $10.04 x 2200 A $10.03 x 400 A $10.02 x 700 A $10.01 x 500 B $10.00 x 300 B $ 9.99 x 200 B $ 9.98 x 1100 B $ 9.97 x 1600 With the above order book, a buyer (for whatever reason) decides to enter into this market a new... He looks quickly at the order book and decides that he likes the current closest ASK price of $10.01 for his 400 shares that he wants to buy and as such, he uses a market order to by 400 shares. The order book now changes to this: A $10.04 x 2200 A $10.03 x 400 A $10.02 x 700 A $10.01 x 100 400 shares were bought by our new buyer from this seller at $10.01 B $10.00 x 300 B $ 9.99 x 200 B $ 9.98 x 1100 B $ 9.97 x 1600 Now, a new buyer comes into the market and he sees that there are only 100 shares being offered for sale at $10.01, but he needs to buy 2000 shares, so what he does is use a limit order to buy 2000 shares at $10.02. The order book would change as follows: A $10.04 x 2200 A $10.03 x 400 B $10.02 x 1200 The 100 shares at $10.01 were "taken out" by this order, along with the 700 shares at $10.02. The remaining 1200 shares this guy wants to purchase stay in the order book at his limit price of $10.02 B $10.00 x 300 B $ 9.99 x 200 B $ 9.98 x 1100 B $ 9.97 x 1600 What happened above? The new buyer just took out the 100 shares that were being offered at $10.01 and took out the other 700 shares that were being offered at $10.02. But because he used a limit order to BUY, the rest of his unexecuted 1200 shares now sit as the current new bid at $10.02. Notice that we left a tiny price point in the book empty ($10.01). This leaves some room for our next buyer who decides that he has 500 shares that he'd (or she) would like to buy, but the current best offer of $10.03 is just a tiny bit too high for him, so he decides to enter a limit order to buy his 500 shares at $10.01. The book changes to this: A $10.04 x 2200 A $10.03 x 400 B $10.02 x 1200 B $10.01 x 500 This guy's 500 shares are added to the book at his limit price of $10.01 B $10.00 x 300 B $ 9.99 x 200 B $ 9.98 x 1100 B $ 9.97 x 1600 Now, just for one more interesting scenario, we have a seller that starts to get anxious. The guy sitting at $10.04 with 2200 decides to say "screw it" and simply sells all of 2200 shares "at market". The order book now looks like this: A $10.03 x 400 B $ 9.98 x 1100 B $ 9.97 x 1600 So, basically, he just took out the $10.02, $10.01, $10.00 and $9.99 bids and left a wide hole in the current best bid/ask. This leaves room for additional buyers and sellers to enter this rather widened market and offer better bids/asks, to hopefully make a better market for new buyers/sellers. Does that make sense as to how markets actually move? This happens all day on thousands of instruments with varying degrees of buyers/sellers with tons and tons of money. This is how prices actually move. In this sense, when you really take a close look at how prices work, you will notice that it becomes less obvious as to what the price of a certain equity actually is... Price becomes a very interesting thing and can be influenced greatly by those with deep, deep pockets.....

Hey There momop, I'm very glad that my posts helped you. That's great news. I wish you tremendous success as you progress on your journey. Jason

Woo Hoo!!!! Congrats Dinero. Jason

Bathrobe, Thanks. I'm glad that my system was able to inspire you in some way. That's way cool..... Jason