6 ms·

I lost about $100k doing algorithmic trading. I spent the better part of 2 years after work immersing myself in algorithmic trading, understanding the architect

I lost about $100k doing algorithmic trading. I spent the better part of 2 years after work immersing myself in algorithmic trading, understanding the architecture of the stock market, and getting very very deep into the topic. I found an algorithm that was wildly positive, and traded it on 3 separate markets every night. I learned a lot and I love everything I learned, but it was a very expensive lesson. The #1 thing I learned was that algo trading is mostly psychological, at least for me. I was making big bets (a few thousand dollars per trade) every night and it was emotionally exhausting, and I couldn't handle the pressure. The worst part is that I didn't trust the algorithm, and would cut the trades short instead of waiting for the full profit (or loss). That messed with my results, and in the end it turned into a disaster. But programming for myself and using real money was such an educational experience. Any little bug meant that I could lose a lot of money so I bug-tested the most I've ever done in my life. And I did things like write my own multi-threaded backtester, working on hundreds of gigabytes of data, so I learned a lot there too. It was a lot of fun, very very expensive fun.

mirceal 8y ago
if you've lost 100k, you had 100k to spare. agree it was an expensive lesson.
- pfarnsworth 8y ago
  Yes, it sucked losing that much money but I'm lucky and grateful that it didn't alter anything about my life.
- icebraining 8y ago
  if you've lost 100k, you had 100k to spare Not if you're leveraging! For an extreme example, this guy lost $210k of borrowed money: https://www.bogleheads.org/forum/viewtopic.php?f=10&t=5934&sid=e650e39b31b9f119b3a8da32242c7049 https://www.bogleheads.org/forum/viewtopic.php?f=10&t=5934&s...
  - downandout 8y ago
    Another example is Chris Sacca, who burned through $16 million - $12 million in paper gains, and $4 million in borrowed money. From [1]: Starting with around $10-$20,000 in what were college loans, Sacca realized that brokers weren’t accounting for margin usage on a real-time basis. The result was that even though firms were only allowing 50% margin, as long as a customer closed a trade before the trade settled, T+3, and showed proper equity in the account, positions larger than 50% margin usage could be opened. Picking winners of stocks that by Sacca’s account had risen 40x, and betting with positions well above his margin requirements, his account grew to $12 million in 18 months. But, as we all know, the record levels of the Nasdaq and the dot com bubble of that time eventually burst....When it did burst, and even though the damage was from holding just two stocks, Sacca found himself in the hole with a $4 million negative balance. [1] https://www.financemagnates.com/forex/brokers/chris-saccathe-4million-negative-balance-salinger-group-twitter/ https://www.financemagnates.com/forex/brokers/chris-saccathe...
chad_strategic 8y ago
Interactive Brokers has a paper trading account. so you can test your trades in real time but not make any money. I couldn't image going into production right away. My algo are good, but they also have some loops that kept buying stock, when it should have stopped. Thankfully it was just paper trading.
- pfarnsworth 8y ago
  I backtested thoroughly and paper traded before going live. Paper trading is nothing like real money trading. That's also one of the first things you learn, it's like a different dimension.
  - tim333 8y ago
    You probably should have taken smaller bets and left the algorithm to it?
ruok0101 8y ago
I wrote a few indicators on Tradestation back in the mid 2000's and had a similar result after using them with an automated trading strategy, albeit less loss (somewhere around $10k). I was trading on 3-1 margin and closed all positions before the end of the day. Very stressful, I too let emotion interrupt trades, etc. Fun to develop, painful to execute. I'm much happier creating startups! =)

mirceal 8y ago

if you've lost 100k, you had 100k to spare. agree it was an expensive lesson.

pfarnsworth 8y ago
Yes, it sucked losing that much money but I'm lucky and grateful that it didn't alter anything about my life.
icebraining 8y ago
if you've lost 100k, you had 100k to spare Not if you're leveraging! For an extreme example, this guy lost $210k of borrowed money: https://www.bogleheads.org/forum/viewtopic.php?f=10&t=5934&sid=e650e39b31b9f119b3a8da32242c7049 https://www.bogleheads.org/forum/viewtopic.php?f=10&t=5934&s...
- downandout 8y ago
  Another example is Chris Sacca, who burned through $16 million - $12 million in paper gains, and $4 million in borrowed money. From [1]: Starting with around $10-$20,000 in what were college loans, Sacca realized that brokers weren’t accounting for margin usage on a real-time basis. The result was that even though firms were only allowing 50% margin, as long as a customer closed a trade before the trade settled, T+3, and showed proper equity in the account, positions larger than 50% margin usage could be opened. Picking winners of stocks that by Sacca’s account had risen 40x, and betting with positions well above his margin requirements, his account grew to $12 million in 18 months. But, as we all know, the record levels of the Nasdaq and the dot com bubble of that time eventually burst....When it did burst, and even though the damage was from holding just two stocks, Sacca found himself in the hole with a $4 million negative balance. [1] https://www.financemagnates.com/forex/brokers/chris-saccathe-4million-negative-balance-salinger-group-twitter/ https://www.financemagnates.com/forex/brokers/chris-saccathe...

chad_strategic 8y ago

Interactive Brokers has a paper trading account. so you can test your trades in real time but not make any money. I couldn't image going into production right away. My algo are good, but they also have some loops that kept buying stock, when it should have stopped. Thankfully it was just paper trading.

pfarnsworth 8y ago
I backtested thoroughly and paper traded before going live. Paper trading is nothing like real money trading. That's also one of the first things you learn, it's like a different dimension.
- tim333 8y ago
  You probably should have taken smaller bets and left the algorithm to it?

ruok0101 8y ago

I wrote a few indicators on Tradestation back in the mid 2000's and had a similar result after using them with an automated trading strategy, albeit less loss (somewhere around $10k). I was trading on 3-1 margin and closed all positions before the end of the day. Very stressful, I too let emotion interrupt trades, etc. Fun to develop, painful to execute. I'm much happier creating startups! =)