Hi caementarius, Thanks for checking our website! At this point, I think we are going to concentrate on supporting traders with a track record. I know it seems unfair to those who want to get into the industry and it is certainly not what we were hoping for at the beginning, but it seems to be a more certain path to make money. The backtest platform might benefit from those improvements, but in general, they will be too specific to some people to be of much benefit to the user of the platform. In particular, we do not plan to work on ETFs in the near future. Sorry. elefevre PS: I want to make clear that, although I am an employee, I am not a decision maker here at Algodeal and I'm contributing to this forum on my own. Please do not hold it against me it things do not turn exactly as I predicted. ;-)
elefevre - Any updates on whether this is the final chosen direction? I am guessing that the crowd didn't show up in the numbers/power needed to make it worth it. If you're going for those with track records, there's not a lot to differentiate algodeal from any number of shops out there. I have some more time now but I don't want to spend much time with it if there's no chance of building a relationship - and if they may just ditch the public interface.
This is how you create an intraday algo trading business: -idea generation -write down your strategy -code your strategy (TradeStation, NinjaTrader, Tradelink, Excel, Matlab, etc) -backtest -walk forward optimization test on multiple instruments (look for profit factor > 2, # trades 2-4 per day, sharpe ratio = 2+) -go over your forward test trades to confirm they make sense -make changes or drop the strategy -demo account forward test -open a small account (with fx start with mini-lots) if your strategy looks good -start building a live track history (3-6 months depending on frequency) -go see a prop firm (equities, futures, fx), show them your track record and that you made some cash -negotiate a deal (it should start around 50/50 for no capital contribution and no salary)
Looks like www.algodeal.com is offline and likely out of business. That's too bad, I liked their idea. Do you think any other companies will crowdsource the process from idea to execution? I think part of AlgoDeal's problem was that they let anyone do it but had not much in the way of contact with system makers - likely envisioning hordes of quants using the platform. My guess is relatively few took the time to check it out.
Yes, the company closed down a few months ago. We have all moved on. Too bad it didn't work out. FWIW, we (some of the former employees) are thinking of buying and open sourcing the code base. Difficult to say whether it will work out.
Kudos to you all for giving it a go. If open-sourced, clients would still need a couple of components: - data - some number of servers to make the cloud optimization worthwhile One increase in efficiency I'm looking to do with my custom backtesting system is to save derived data, so that backtests are more about just trade logic. For example, If I'm calculating VWAP for a certain timeframe on some security, I'll save the derived data - those are calculations that really only have to be made once and are thence available just by lookup.