3 ms·

I would love to hear about some recommended reading and other sources which could help hobby traders like me to develop their own algorithms.

Kranar 5y ago
Good sources are very hard to find because most of it is absolute trash intended to appeal to a certain audience who think there's money to be made off of reverse Fibonacci patterns or other silly sounding technical indicators that once again, sound technical and fancy but are completely useless. If you want to know what it's really like to be a quant, review stuff from the ARPM; everyone I hire goes through their 6 day bootcamp but they have other materials as well: https://www.arpm.co/ https://www.arpm.co/ And as for books, Algorithmic and High Frequency Trading covers the foundations: https://www.amazon.com/Algorithmic-High-Frequency-Trading-%C3%81lvaro-Cartea/dp/1107091144 https://www.amazon.com/Algorithmic-High-Frequency-Trading-%C... Those are pretty good sources to get an overview of what actual quants at successful firms know. Quantitative trading is technical, I don't want to give the impression that it's not technical... but it's not "fancy" technical. It's more along the lines of rigorous and iron clad instead of flashy and sophisticated. Every strategy is built up step by meticulous step in precise detail and every step needs to be rigorously justified and experimentally verified. Generally the thought process starts from the assumption that there is no money to be made on the stock market, either due to perfect efficiency or things like fees eating up any potential profits... when we talk about models, the models we construct describe how the stock market would behave if it were perfectly efficient, ie. free of any arbitrage opportunity. Then given our model of a perfectly efficient stock market, we simulate what we should expect to observe in such a perfectly efficient market... we then investigate empirically whether these observations happen in reality. Is the market genuinely efficient all day every day across every security. For some phenomenon it really is, but sometimes the market deviates from the model, so our model is either incorrect or an arbitrage opportunity has presented itself. If an arbitrage opportunity presents itself, we investigate how feasible it is to capture it, things like engineering effort, risk factors, profitability etc... If all of that works out, then we get to work constructing a state machine for an algorithm to capture that opportunity. We implement the state machine, write tests for it, run it through our backtester, then run it through our live simulator, and after everything checks out we deploy it live. Every algorithm is treated like a person, it's given its own human-like name, has its own account, its own set of permissions, capital allocated to it, risk profile, and algorithms are evaluated on a daily basis to reallocate capital to them and modify their risk profile.
- nradov 5y ago
  Fascinating. Could you describe how you deconflict your various named algorithms and prevent them from competing with each other?
  - sampo 5y ago
    The description says they already have a market simulator, which must be a significant piece of software to initially construct. Once you have that, you could just run simulations and observe empirically, if your algorithms end up competing against each others. Not that I know, how they actually do it.
  - Kranar 5y ago
    We don't stop algos from competing with one another, that said it is illegal for us as a firm to be both sides of a trade, that is we are buying/selling to ourselves in what is referred to as a wash trade. This can happen if algo A is buying Microsoft at the same time that algo B is selling Microsoft. We do prevent this by using a fairly standard component called an internalizer. Almost all of our orders go through such an internalizer and if it sees that two orders are matching one another, the internalizer will perform that match directly instead of having it go to the market. This happens quite frequently in fact, and it's good to be able to catch it beforehand, avoids paying fees, having any kind of market impact, as well as obviously complying with regulations.
- jedimind 5y ago
  Thank you so much for taking the time and sharing your knowledge, much appreciated!
- mrjangles 5y ago
  Assuming one has an appropriate PhD in math/physics, what other kind of general knowledge questions would you be asking people in an interview? For example, I've read that one should learn modern portfolio theory before going for an interview. Is this sort of thing true, or would that be a waste of time?
  - 5y ago

Kranar 5y ago

Good sources are very hard to find because most of it is absolute trash intended to appeal to a certain audience who think there's money to be made off of reverse Fibonacci patterns or other silly sounding technical indicators that once again, sound technical and fancy but are completely useless. If you want to know what it's really like to be a quant, review stuff from the ARPM; everyone I hire goes through their 6 day bootcamp but they have other materials as well: https://www.arpm.co/ https://www.arpm.co/ And as for books, Algorithmic and High Frequency Trading covers the foundations: https://www.amazon.com/Algorithmic-High-Frequency-Trading-%C3%81lvaro-Cartea/dp/1107091144 https://www.amazon.com/Algorithmic-High-Frequency-Trading-%C... Those are pretty good sources to get an overview of what actual quants at successful firms know. Quantitative trading is technical, I don't want to give the impression that it's not technical... but it's not "fancy" technical. It's more along the lines of rigorous and iron clad instead of flashy and sophisticated. Every strategy is built up step by meticulous step in precise detail and every step needs to be rigorously justified and experimentally verified. Generally the thought process starts from the assumption that there is no money to be made on the stock market, either due to perfect efficiency or things like fees eating up any potential profits... when we talk about models, the models we construct describe how the stock market would behave if it were perfectly efficient, ie. free of any arbitrage opportunity. Then given our model of a perfectly efficient stock market, we simulate what we should expect to observe in such a perfectly efficient market... we then investigate empirically whether these observations happen in reality. Is the market genuinely efficient all day every day across every security. For some phenomenon it really is, but sometimes the market deviates from the model, so our model is either incorrect or an arbitrage opportunity has presented itself. If an arbitrage opportunity presents itself, we investigate how feasible it is to capture it, things like engineering effort, risk factors, profitability etc... If all of that works out, then we get to work constructing a state machine for an algorithm to capture that opportunity. We implement the state machine, write tests for it, run it through our backtester, then run it through our live simulator, and after everything checks out we deploy it live. Every algorithm is treated like a person, it's given its own human-like name, has its own account, its own set of permissions, capital allocated to it, risk profile, and algorithms are evaluated on a daily basis to reallocate capital to them and modify their risk profile.

nradov 5y ago
Fascinating. Could you describe how you deconflict your various named algorithms and prevent them from competing with each other?
- sampo 5y ago
  The description says they already have a market simulator, which must be a significant piece of software to initially construct. Once you have that, you could just run simulations and observe empirically, if your algorithms end up competing against each others. Not that I know, how they actually do it.
- Kranar 5y ago
  We don't stop algos from competing with one another, that said it is illegal for us as a firm to be both sides of a trade, that is we are buying/selling to ourselves in what is referred to as a wash trade. This can happen if algo A is buying Microsoft at the same time that algo B is selling Microsoft. We do prevent this by using a fairly standard component called an internalizer. Almost all of our orders go through such an internalizer and if it sees that two orders are matching one another, the internalizer will perform that match directly instead of having it go to the market. This happens quite frequently in fact, and it's good to be able to catch it beforehand, avoids paying fees, having any kind of market impact, as well as obviously complying with regulations.
jedimind 5y ago
Thank you so much for taking the time and sharing your knowledge, much appreciated!
mrjangles 5y ago
Assuming one has an appropriate PhD in math/physics, what other kind of general knowledge questions would you be asking people in an interview? For example, I've read that one should learn modern portfolio theory before going for an interview. Is this sort of thing true, or would that be a waste of time?
- 5y ago