Backtesting / virtual portfolio
Dear Team, thanks for your replies (despite not having raised a typical R-issue) - but I guess it is of interest for many :-) There is no doubt that information is expensive - especially on stock quotes. However, I would not judge / base the ability of someone only with respect to the amount of money he spends on real-time data :-) Just owning an expensive Bloomberg terminal does not help - and a first step to increase the net return is decrease costs such as trading costs, etc. I mean the main problem is: just because I am spending a lot of money (possibly thousands of Euro)it does not automatically mean, that I am allowed to share the pricing information publicly or somebody else - which is the main problem in my case. The next thing is that I do not want to spend thousands of Euros (or dollars) just to have a "virtual portfolio" I could share - I mean, honestly, what is the point in doing so? I found stocktrak, warketwatch and some others would offer the benefits for most of my purposes more or less for free - but they don't offer pricing of futures in German indices. It does add some more value if you can have a "neutral quotes and trading" platform to test your strategies - I never really trsuted backtestingstrategies. So - I am still open for ideas with respect to that :-) Thanks a lot for all of your help! I really appreciate it! I searched for that for quite a while - but it seems like there is no possibility to show "what would happen to a portfolio" in realtime to potentially interested persons - unless you have a fund or do it with real money... :-/ I only carry out trades on a day-to-day basis, so no intraday trading... Lui
On 11/29/11, Mark Leeds <markleeds2 at gmail.com> wrote:
hi: brian says "not planning on making hundreds of trades per day" because if you're doing hundreds of trades per day, then this means you have relatively short term holding periods, probably on the order of minutes. so, it's hard to know if you're gonna be able to get in to the trade at the time and price you want and out of the trade at the time and price you want. ( when you backtest or paper trade, you're seeing either trades or quotes and you can't be sure you can DEFINITELY be sure that your trade can be filled or closed out at those prices ) and then, since the trades have short term holding periods, this ( whether you can get in or can't get in or get out or can't get out ) will effect the results a lot. on the other hand, in a scenario where you're holding periods are longer, it doesn't matter as much what price and time you get in. for example, if you're holding period is say a week or more, then it's probably not going to make a heck of a difference whether you get into a trade mid-day or end of day or get out mid-day or end of day. so, the paper trading will be more realistic in the longer holding period scenario. I hope that clarifies what you didn't understand. mark On Mon, Nov 28, 2011 at 11:18 PM, Christofer Bogaso < bogaso.christofer at gmail.com> wrote:
Sorry for my ignorance............" are not planning on making hundreds of trades per day, " why it is so? Thanks and regards, -----Original Message----- From: r-sig-finance-bounces at r-project.org [mailto:r-sig-finance-bounces at r-project.org] On Behalf Of Brian G. Peterson Sent: Tuesday, November 29, 2011 7:43 AM To: Lui ## Cc: r-sig-finance Subject: Re: [R-SIG-Finance] Backtesting / virtual portfolio On Tue, 2011-11-29 at 01:00 +0100, Lui ## wrote:
after backtesting strategies for a while I am wondering how to actually "prove" they "kind a work" in a "as realistic as possible" market environment. Even though I know this is not exactly related to R - it is just one step further. Do you have any suggestions for Portfolio Simulations that could be shared with others online - and enable a real time (or close to real time) portfolio tracking? I came across www.stocktrak.com but I do not know whether that is the right thing for me - especially since I am mainly interested in German Future contracts and $60 per month seem quite expensive... The actual underlying issue is "building up some credibility" with respect to the buy/sell suggestions a certain trading strategy may suggest - I don't think my bank account statement will serve that purpose... I was looking for something more transparent that could be incorporated into a website... Thanks a lot for your suggestions and help - and again my apologies - its not directly R, but closely linked to the topic... :-) Lui
Definitely not R, but here are a couple thoughts: Many brokers, including InterActiveBrokers, support paper trading, where you can trade in real time and keep track of P&L. If you place realistic orders (limit orders of reasonable size), then you can assume that your paper trading, if done in real time, closely mimics your potential results on a live account during that period. I suspect that if $60/month is too much to spend for a useful tool, you need to consider whether you are ready to be a professional trader. We spend thousands of dollar per month just on data, and thousands more on trading infrastructure, for a very small trading operation. If you are making realistic assumptions and are not planning on making hundreds of trades per day, you should be able to get more than 90% correlations between backtests and production in R. Regards, - Brian -- Brian G. Peterson http://braverock.com/brian/ Ph: 773-459-4973 IM: bgpbraverock
_______________________________________________ R-SIG-Finance at r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-sig-finance -- Subscriber-posting only. If you want to post, subscribe first. -- Also note that this is not the r-help list where general R questions should go. _______________________________________________ R-SIG-Finance at r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-sig-finance -- Subscriber-posting only. If you want to post, subscribe first. -- Also note that this is not the r-help list where general R questions should go.