5 ms·

What you have built looks great and I am sure a lot of work went into it. However, as someone with a bit of trading experience (running a profitable custom-buil

What you have built looks great and I am sure a lot of work went into it. However, as someone with a bit of trading experience (running a profitable custom-built trading system myself) who has friends that lost money I would like to post a few words of warning to anyone who believes they can build profitable trading strategies on top of someone else's platform: It's extremely unlikely. While you may hear stories of people making money, most of them are pure luck (I recommend the book Fooled By Randomness). To make a stable incoming with trading you must have a consistent "edge", a competitive advantage that other traders don't. Possibilities here include 1. data, that could be be cleaner, more fine grained (L2/L3 book data, better reconstructed, etc) 2. infrastructure. This includes highly latency-optimized server placement, custom API integrations, fault tolerance, dealing with api issues, etc 3. "smarter" trading strategies - What many companies are selling you is that you can make quick $$$ by coming up with some secret trading strategy (3). That's how they make money off you. These charting patterns are pseudoscience. 99% of all edge in crypto trading is in #1 and #2 - infrastructure and data, with relatively simple well-known strategy algorithms. By "outsourcing" this to a platform you are giving up your edge and set yourself up to lose money in the long run. Also, ask yourself why someone would offer a platform to build profitable strategies instead of simply trading themselves based on their competitive advantage. The answer almost always is: Because they failed to trade profitably and pivoted to selling their (unprofitable) infra. I don't want to put down what you have built. I know firsthand how hard it is to build some of this infrastructure. I would just like to warn people to not easily trust trading infra providers before they lose money.

rohitgoyal 8y ago
What you are saying is right specially for HFT and also because no platform have been able to achieve that till now. We decided to build this only after talking to 100s of traders personally. We definitely can’t help people develop profitable strategies but if someone has a strategy, she wouldn’t be stuck because of resources/capital limitation. Having said that we are going to keep building the state of the art infrastructure to support more efficient trading as well. Thanks for taking so much time to provide feedback.
- lquist 8y ago
  I don’t think this really addresses his criticisms of your startup...
- curiousjorge 8y ago
  that has nothing to do with what hes saying
- racecar789 8y ago
  Threads are going off the rails a bit. You built a fine tool for others to use. With a free tier. Nice work...
  - rohitgoyal 8y ago
    Thanks for your support. :)
    - curiousjorge 8y ago
      Look...we can see what you are doing...we are not stupid, you are completely side stepping the important issues. A lot of us work on wall street and we can smell bullshit from a mile away. And this smells like a pile of horse shit.
pault 8y ago
Totally naive question: what if you aren't interested in HFT but just want to take some of the manual labor out of making less frequent trades? Is that any less reliable than sitting at your computer watching the order books or do the exchanges already have all the tools you would need for that?
- throwawaykyok 8y ago
  Yes, that's a valid use case for this tool. Exchanges typically don't have this functionality. My opinion based on my personal experience is that it is rather unlikely to build profitable algorithms based purely on "technical analysis patterns": (1) Most of these patterns are pseudoscience, you can google for more info, only a few have even a bit of science behind them (2) There isn't enough data to train a model on minutely or hourly data. The crypto markets change extremely quickly (data distribution shift) so that most of the data you would be using to optimize model parameters is outdated and leads to false conclusions. To make money off less frequent trades I believe (again, just my opinion) that your decisions must be based on news/insight/insider info, not charting patterns. In other words, more fundamentals. And that's much harder to automate. Of course, you will find people that tell you the opposite and that they make money. My response to this would be: Of course there are those that make money due to simple laws of probability. With a lot of people trading there will be some winners. But it is mostly due to luck rather than skill. And the fewer trades your make (lower frequency) the harder it is to assign any kind of significance numbers to the results.
  - luxpir 8y ago
    It doesn't sound like you've thought this through. Technical analysis isn't typically based on patterns. Banks trade using TA (see Lehman Bros Forex trading handbook - available online) - it's not strange or unusual to trade a mean reversion strategy at HFT or retail trading levels. Nor a momentum strategy, or even trend following. Nobody serious trades the candlestick patterns you're referring to in algo trading. There is enough data to train all forex pairs - tick data going back decades, for free, via Dukascopy and other sources. You're spreading FUD, and while there is definitely a need for people to be cautious, your warnings are not based in reality. Better to warn them about the 70-90% losing rate of all retail traders (reported officially by all UK brokers), or the perils of overleveraging, or the risk of gapping/illiquidity in crypto... I could go on. But not what you mention - that's a strawman.
nightski 8y ago
In response to your second point - it's largely because it takes quite a bit of capital to build wealth via trading. Much more capital than is needed starting a SaaS company. It's hard to win picking individual stocks, and yes you should not only hold a few stocks at a time with all of your portfolio. But you don't have to bet your entire portfolio. You can only trade a portion of it hedging your risk. In addition, while it may be more of a fools game to go big on one or two stocks, you can look at your risk profile and align your long term strategy with different sectors and tax strategies. I am no means a trading expert, but to discourage people from investing just because certain types of trading is more akin to playing the lottery - is just as irresponsible. People should be encouraged to be more engaged with their money and at the very least follow how their index fund is doing, how it's allocated, and learn why it is allocated the way it is.
- pault 8y ago
  I think GP is cautioning against high frequency algorithmic trading, not investing in general.
throwawaymath 8y ago
Out of curiosity, do you also have experience trading equities/derivatives? I largely agree with you (especially the words of caution and about buying other peoples' trading infra), though I'd push back a bit on your assessment of automated trading at lower frequencies than HFT. I'm in strong agreement with you about the lack of utility for what is typically called "charting" or "technical analysis", but I don't think it would be fair to characterize all of stat arb at less than HFT timescales this way. But my experience is entirely outside cryptocurrencies, so take that for what you will. I'm just curious to hear if you're also applying this perspective outside cryptocurrency trading. If anything I'd expect it to be far easier to profitably trade cryptocurrencies these days since the space is still so inefficient.
- throwawaykyok 8y ago
  I have a bit of experience trading derivatives, but little compared to crypto. Everything I said was specific to crypto only. A crucial distinction between crypto and equities is direct market access. Everyone in crypto gets DMA and API access with a few clicks only. Most of the low-hanging arb fruit is thus already taken or getting harder to take and moving to higher frequencies. There are still low-frequency arb opportunities in crypto, but for many of them the edge is in efficient/safe fiat currency movement across country orders, i.e. having the right citizenships and bank accounts and clearances. In many countries it's not easy to open bank accounts for crypto trading these days. People often think there are arb opportunities in crypto when there are none because the price already includes the inefficiencies, latencies, and difficulties of moving fiat across country borders or taking money out. Also, the price often includes the risk of the exchange being hacked or running away with your money, which has happened a lot recently. So when people look at arb opportunities they often don't take into account that these risks must be reflected in the exchange prices. Sketchy exchange prices are lower due to the risk factor of having balances there. That's not the case for regulated financial markets.
  - bitcoinmoney 8y ago
    Just curious how much money you made in total and you’re sharpe ratio? What about win rate and how man trades you made? We could use that to determine if you have an edge (not saying you don’t).
eternalny1 8y ago
> Also, ask yourself why someone would offer a platform to build profitable strategies instead of simply trading themselves based on their competitive advantage. The answer almost always is: Because they failed to trade profitably and pivoted to selling their (unprofitable) infra. This is correct. It's so obvious but most people simply don't see it for what it is. The interesting thing about these platforms is that they can actually record what OTHER people are doing, see what the successful people do, and possibly just copy them.
- creatornator 8y ago
  I'm really glad these points are being pushed in this thread--it is almost impossible to tune a portfolio for performance without knowing something the market doesn't (data, computing power, information). That's why I only ever tune my portfolio for the level of risk that I can tolerate. Because that's pretty much the only thing you _can_ tune for.
sprash 8y ago
I disagree with the notion that your edge has to rely mainly on the infrastructure or insider knowledge. In today's markets you often have the case that very wealthy and high volume market participants have to make ends meet with rather low liquidity markets. This leaves more than enough room for small fish to arbitrate since the price allocation is flawed.

rohitgoyal 8y ago

What you are saying is right specially for HFT and also because no platform have been able to achieve that till now. We decided to build this only after talking to 100s of traders personally. We definitely can’t help people develop profitable strategies but if someone has a strategy, she wouldn’t be stuck because of resources/capital limitation. Having said that we are going to keep building the state of the art infrastructure to support more efficient trading as well. Thanks for taking so much time to provide feedback.

lquist 8y ago
I don’t think this really addresses his criticisms of your startup...
curiousjorge 8y ago
that has nothing to do with what hes saying
racecar789 8y ago
Threads are going off the rails a bit. You built a fine tool for others to use. With a free tier. Nice work...
- rohitgoyal 8y ago
  Thanks for your support. :)
  - curiousjorge 8y ago
    Look...we can see what you are doing...we are not stupid, you are completely side stepping the important issues. A lot of us work on wall street and we can smell bullshit from a mile away. And this smells like a pile of horse shit.

pault 8y ago

Totally naive question: what if you aren't interested in HFT but just want to take some of the manual labor out of making less frequent trades? Is that any less reliable than sitting at your computer watching the order books or do the exchanges already have all the tools you would need for that?

throwawaykyok 8y ago
Yes, that's a valid use case for this tool. Exchanges typically don't have this functionality. My opinion based on my personal experience is that it is rather unlikely to build profitable algorithms based purely on "technical analysis patterns": (1) Most of these patterns are pseudoscience, you can google for more info, only a few have even a bit of science behind them (2) There isn't enough data to train a model on minutely or hourly data. The crypto markets change extremely quickly (data distribution shift) so that most of the data you would be using to optimize model parameters is outdated and leads to false conclusions. To make money off less frequent trades I believe (again, just my opinion) that your decisions must be based on news/insight/insider info, not charting patterns. In other words, more fundamentals. And that's much harder to automate. Of course, you will find people that tell you the opposite and that they make money. My response to this would be: Of course there are those that make money due to simple laws of probability. With a lot of people trading there will be some winners. But it is mostly due to luck rather than skill. And the fewer trades your make (lower frequency) the harder it is to assign any kind of significance numbers to the results.
- luxpir 8y ago
  It doesn't sound like you've thought this through. Technical analysis isn't typically based on patterns. Banks trade using TA (see Lehman Bros Forex trading handbook - available online) - it's not strange or unusual to trade a mean reversion strategy at HFT or retail trading levels. Nor a momentum strategy, or even trend following. Nobody serious trades the candlestick patterns you're referring to in algo trading. There is enough data to train all forex pairs - tick data going back decades, for free, via Dukascopy and other sources. You're spreading FUD, and while there is definitely a need for people to be cautious, your warnings are not based in reality. Better to warn them about the 70-90% losing rate of all retail traders (reported officially by all UK brokers), or the perils of overleveraging, or the risk of gapping/illiquidity in crypto... I could go on. But not what you mention - that's a strawman.

nightski 8y ago

In response to your second point - it's largely because it takes quite a bit of capital to build wealth via trading. Much more capital than is needed starting a SaaS company. It's hard to win picking individual stocks, and yes you should not only hold a few stocks at a time with all of your portfolio. But you don't have to bet your entire portfolio. You can only trade a portion of it hedging your risk. In addition, while it may be more of a fools game to go big on one or two stocks, you can look at your risk profile and align your long term strategy with different sectors and tax strategies. I am no means a trading expert, but to discourage people from investing just because certain types of trading is more akin to playing the lottery - is just as irresponsible. People should be encouraged to be more engaged with their money and at the very least follow how their index fund is doing, how it's allocated, and learn why it is allocated the way it is.

pault 8y ago
I think GP is cautioning against high frequency algorithmic trading, not investing in general.

throwawaymath 8y ago

Out of curiosity, do you also have experience trading equities/derivatives? I largely agree with you (especially the words of caution and about buying other peoples' trading infra), though I'd push back a bit on your assessment of automated trading at lower frequencies than HFT. I'm in strong agreement with you about the lack of utility for what is typically called "charting" or "technical analysis", but I don't think it would be fair to characterize all of stat arb at less than HFT timescales this way. But my experience is entirely outside cryptocurrencies, so take that for what you will. I'm just curious to hear if you're also applying this perspective outside cryptocurrency trading. If anything I'd expect it to be far easier to profitably trade cryptocurrencies these days since the space is still so inefficient.

throwawaykyok 8y ago
I have a bit of experience trading derivatives, but little compared to crypto. Everything I said was specific to crypto only. A crucial distinction between crypto and equities is direct market access. Everyone in crypto gets DMA and API access with a few clicks only. Most of the low-hanging arb fruit is thus already taken or getting harder to take and moving to higher frequencies. There are still low-frequency arb opportunities in crypto, but for many of them the edge is in efficient/safe fiat currency movement across country orders, i.e. having the right citizenships and bank accounts and clearances. In many countries it's not easy to open bank accounts for crypto trading these days. People often think there are arb opportunities in crypto when there are none because the price already includes the inefficiencies, latencies, and difficulties of moving fiat across country borders or taking money out. Also, the price often includes the risk of the exchange being hacked or running away with your money, which has happened a lot recently. So when people look at arb opportunities they often don't take into account that these risks must be reflected in the exchange prices. Sketchy exchange prices are lower due to the risk factor of having balances there. That's not the case for regulated financial markets.
- bitcoinmoney 8y ago
  Just curious how much money you made in total and you’re sharpe ratio? What about win rate and how man trades you made? We could use that to determine if you have an edge (not saying you don’t).

eternalny1 8y ago

> Also, ask yourself why someone would offer a platform to build profitable strategies instead of simply trading themselves based on their competitive advantage. The answer almost always is: Because they failed to trade profitably and pivoted to selling their (unprofitable) infra. This is correct. It's so obvious but most people simply don't see it for what it is. The interesting thing about these platforms is that they can actually record what OTHER people are doing, see what the successful people do, and possibly just copy them.

creatornator 8y ago
I'm really glad these points are being pushed in this thread--it is almost impossible to tune a portfolio for performance without knowing something the market doesn't (data, computing power, information). That's why I only ever tune my portfolio for the level of risk that I can tolerate. Because that's pretty much the only thing you _can_ tune for.

sprash 8y ago

I disagree with the notion that your edge has to rely mainly on the infrastructure or insider knowledge. In today's markets you often have the case that very wealthy and high volume market participants have to make ends meet with rather low liquidity markets. This leaves more than enough room for small fish to arbitrate since the price allocation is flawed.