tag:blogger.com,1999:blog-35364652.post4442813725749998048..comments2024-03-22T10:29:59.088-04:00Comments on Quantitative Trading: What is the probability of profit of your next trade? (Introducing PredictNow.Ai)Ernie Chanhttp://www.blogger.com/profile/02747099358519893177noreply@blogger.comBlogger31125tag:blogger.com,1999:blog-35364652.post-33151162612006318522023-06-08T05:44:37.444-04:002023-06-08T05:44:37.444-04:00Hi,
I don't think positions as features would ...Hi,<br />I don't think positions as features would help, but recent performance of your strategy might.<br />ErnieErnie Chanhttps://www.blogger.com/profile/02747099358519893177noreply@blogger.comtag:blogger.com,1999:blog-35364652.post-64113721668726456642023-06-08T05:05:11.249-04:002023-06-08T05:05:11.249-04:00Hi Ernie
Will including the strategies positions ...Hi Ernie<br /><br />Will including the strategies positions as a feature worsen or improve the algorithms performance?<br /><br />ThanksAnonymousnoreply@blogger.comtag:blogger.com,1999:blog-35364652.post-40077948626299256962021-03-16T21:20:17.547-04:002021-03-16T21:20:17.547-04:00Hi James,
In theory, that would work. In practice,...Hi James,<br />In theory, that would work. In practice, it is extraordinarily hard to generate realistic time series with all the quirks of a real financial time series. So parameters optimized that way may not be suitable for real life.<br /><br />Using our Conditional Parameter Optimization technique, we learn from real time series, but with many combinations of parameters to simulate different backtest outcomes. This technique is made possible by Predictnow.ai's API, and I am happy to disclose more privately!<br /><br />Best,<br />ErnieErnie Chanhttps://www.blogger.com/profile/02747099358519893177noreply@blogger.comtag:blogger.com,1999:blog-35364652.post-62412473302592446262021-03-16T18:35:52.200-04:002021-03-16T18:35:52.200-04:00Hi Ernie,
Thanks for your response, I'll cert...Hi Ernie,<br /><br />Thanks for your response, I'll certainly get in touch once I'm closer to being able to incorporate PredictNow into my framework.<br />In the meantime I figured I'd ask my other questions here for the benefit of other readers.<br /><br />A few years ago you proposed generating synthetic data on which to backtest strategies to correct for overfitting. The strategy parameters that perform well on the synthetic data are the ones that should be selected for live trading. I'm wondering if such a procedure still has a purpose after consideration of a trade outcome classifier, like PredictNow.<br /><br />Say, for example, I have three parameters to fit for my strategy, and each can vary across 10 levels. That gives me 10x10x10=1000 possible parameter combinations for my strategy. I could generate 100,000 synthetic price series on which to test each of the parameter combinations, performing 100 million backtests in the process. I would then keep all parameter combinations that perform well, on average, across those 100,000 synthetic price series, discarding the rest.<br /><br />Say I were to instead train the trade outcome classifier on trades placed by each of those 1000 parameter combinations. Would synthetic data backtesting continue to add value for assessing overfitting? Presumably the trade outcome classifier would be able to determine which of those parameter combinations overfit the data, giving them consistently low probability forecasts as a result and limiting their influence on the 1000 system trade signals. Does synthetic data backtesting serve any purpose in the presence of a trade outcome classifier like PredictNow?<br /><br />Regards,<br /><br />JamesJamesnoreply@blogger.comtag:blogger.com,1999:blog-35364652.post-11524260575725742642021-03-15T13:23:13.736-04:002021-03-15T13:23:13.736-04:00Hi James,
Thanks for your kind words!
You don'...Hi James,<br />Thanks for your kind words!<br /><br />You don't have to predict probability on a per trade basis. You can predict them on a daily basis.<br /><br />Are you trading just one symbol? Or do you want to allocate across different symbols?<br /><br />If it works better for you, please email me (ernest@predictnow.ai) and we can set up a call to discuss.<br /><br />Best,<br />ErnieErnie Chanhttps://www.blogger.com/profile/02747099358519893177noreply@blogger.comtag:blogger.com,1999:blog-35364652.post-46803653440115576862021-03-14T17:51:39.548-04:002021-03-14T17:51:39.548-04:00Hi Ernie,
This is a great piece, and very timely ...Hi Ernie,<br /><br />This is a great piece, and very timely for me since I've just begun to consider how ML can augment my deterministic trading strategies.<br /><br />I have a strategy framework where lots of trading system signals are aggregated at the symbol level to arrive at a final allocation that varies slowly over time. For example, AUDCAD may have an allocation of 0.10 one minute (indicating a 10% long allocation), and an allocation of 0.15 the next minute, a net change of 5% long. It can also turn negative, so the allocation may change from 0.10 to -0.25, indicating a 25% short allocation of capital, and a change of 35%.<br /><br />How could I use PredictNow to improve my trading of the aggregate signal on whole symbols? Due to the features I intend to use, I'm unable to use PredictNow to forecast the probability of success of the individual trades whose signals (short, neutral, long i.e. -1/0/1) contribute to the aggregated symbol allocation. So rather than a single discrete trade whose outcome is well-defined, all I have to go on is the net signal applied to the symbol which varies over time.<br /><br />Do you have any recommendations on how I could use PredictNow on this aggregate signal?<br /><br />Regards,<br /><br />JamesJamesnoreply@blogger.comtag:blogger.com,1999:blog-35364652.post-22073347658527128442020-10-05T15:08:24.892-04:002020-10-05T15:08:24.892-04:00Hi Carlos,
Yes, if you plan to use stop loss and p...Hi Carlos,<br />Yes, if you plan to use stop loss and profit target, you need to use triple barrier labels. Otherwise, you can apply two, or even just one, barrier.<br />ErnieErnie Chanhttps://www.blogger.com/profile/02747099358519893177noreply@blogger.comtag:blogger.com,1999:blog-35364652.post-59552113756612354852020-10-05T15:01:20.777-04:002020-10-05T15:01:20.777-04:00Thanks Ernie. Got it. Do I then need to create the...Thanks Ernie. Got it. Do I then need to create the labels using Lopez de Prado's triple barrier method or is that not a must with your service?Carlosnoreply@blogger.comtag:blogger.com,1999:blog-35364652.post-6806493266268012032020-10-05T15:00:20.050-04:002020-10-05T15:00:20.050-04:00Thanks Ernie. Got it. Do I still need to apply the...Thanks Ernie. Got it. Do I still need to apply the triple barrier-method as mentioned by Lopez de Prado? or is that not necessary to create proper labels?Carlosnoreply@blogger.comtag:blogger.com,1999:blog-35364652.post-77702386125067354912020-10-05T10:51:46.488-04:002020-10-05T10:51:46.488-04:00Hi Carlos,
Currently, you will have to create your...Hi Carlos,<br />Currently, you will have to create your own labels, as we do not know what your stop loss and profit target are.<br /><br />The labels for 0 and 1 are generally for a round trip trade. If you have a portfolio with many stocks, you may instead want to predict the daily returns of the portfolio. That way, you can adjust the capital of the portfolio instead of a single constituent.<br /><br />ErnieErnie Chanhttps://www.blogger.com/profile/02747099358519893177noreply@blogger.comtag:blogger.com,1999:blog-35364652.post-87075385376837206562020-10-05T10:44:01.916-04:002020-10-05T10:44:01.916-04:00Hi Ernie,
Not sure my previous comment made it. I...Hi Ernie,<br /><br />Not sure my previous comment made it. I wanted to know if for the labels I need to apply the triple barrier method before uploading it to PredictNow.ai or it is something that the service does automatically?<br /><br />Also I wasn't sure if for trades that are long for multiple days or weeks, if my indicators are daily, which categories should these labels be if 1 is for profit and 0 is for loss?<br /><br />Thanks,<br /><br />CarlosCarlosnoreply@blogger.comtag:blogger.com,1999:blog-35364652.post-23555195251694024232020-10-05T07:26:38.175-04:002020-10-05T07:26:38.175-04:00Hi,
Yes, if you want to limit your liability in ca...Hi,<br />Yes, if you want to limit your liability in case of Black Swan loss, you should incorporate.<br />To incorporate a LLC is quite trivial if you intend to be the sole member. Just go to bizfilings.com.<br />ErnieErnie Chanhttps://www.blogger.com/profile/02747099358519893177noreply@blogger.comtag:blogger.com,1999:blog-35364652.post-78994091566618060342020-10-04T18:57:27.655-04:002020-10-04T18:57:27.655-04:00Hi Ernie,
Do you have any recommendation on how m...Hi Ernie,<br /><br />Do you have any recommendation on how might one incorporate an LLC to benefit from active trader status? I am working full time, but also execute more than 80-100 trades a month in a personal trading account. I am thinking of eventually quitting my job and trade full time. Should I register as an LLC ?<br /><br />ThanksAnonymousnoreply@blogger.comtag:blogger.com,1999:blog-35364652.post-32535209890684005062020-10-04T06:16:37.548-04:002020-10-04T06:16:37.548-04:00Hi Ernie,
I am confused with the label part. As I...Hi Ernie,<br /><br />I am confused with the label part. As I understood, I can enter in the label my discretionary trades. So 0 when i made no profit and 1 when I made a profit.<br /><br />How do i handle days where no trades (entry or exit) happened? for example say I use the SMA_10 as a features, the SMA_10 has 100 rows (i.e: 100 days lookback). Within these 100 days, say I traded half of the time. Would i just drop the days in the SMA_10 columns where no trades occurred?<br /><br />2nd question related to the first one, Marcos Lopez de Prado talks about the triple-barrier method to label the trades. Am I supposed to label the data using this method or is the service doing this?<br /><br />Thanks for the consideration,<br /><br />CarlosCarlosnoreply@blogger.comtag:blogger.com,1999:blog-35364652.post-57285876722799399622020-09-15T11:26:41.982-04:002020-09-15T11:26:41.982-04:00Hi, I am posting this comment with the permission ...Hi, I am posting this comment with the permission of Ernie to introduce my work called yMTa and hopefully have a broader audience for it's recognition. Thanks Ernie.<br /><br />yMTa, is a freely available Windows software and includes many Market Trading related tools.<br /><br />I am a developer and have implemented many trading algorithms for the past decade and always looking to find professional traders and include their working trading algorithm in yMTa software for free.<br /><br />Please note that, I am not looking for a job or fund and just want you take a look at my work in yMTa or let me have an opportunity to show what can be done using yMTa. I also could be able to add new tools or algorithms to yMTa based on your advise or your more specific requirements.<br /><br />Right now I am looking to find professional people in the field and be able to demonstrate what are the benefits of the tools in yMTa and find ways to make it more useful and have the right path in adding new tools to it.<br /><br />Please take a look at my software and let's talk if you are interested.<br /><br />yMTa, <br />A freely available Windows software, Includes many Market Trading related tools.<br />t.me/yMTaCh<br />Yashil MultiTrding<br />https://t.me/yMTaCh<br /><br />Email: yashil1@yahoo.com<br />Telegram: @Yashil1<br /><br />yMTa Download:<br />https://drive.google.com/drive/folders/1BHzFSr2pCve2-3KJCRkVOBjYXxaubFx9<br /><br />yMTa English Telegram channel: t.me/yMTaCh<br /><br />Old yMTa Persian Telegram channels:<br />https://t.me/yMT4ch<br />https://t.me/yMT4notify<br />https://t.me/yMT4demo<br />Yashil1https://www.blogger.com/profile/02763770897327874692noreply@blogger.comtag:blogger.com,1999:blog-35364652.post-72611803958054867692020-08-14T14:24:19.189-04:002020-08-14T14:24:19.189-04:00Hi Ankit,
You can apply it to any frequency. It al...Hi Ankit,<br />You can apply it to any frequency. It all depends on your labels. Labels can be 1 minute returns, or 1 month returns.<br /><br />Yes, we can take binary, or in general categorical, features.<br /><br />Best,<br />ErnieErnie Chanhttps://www.blogger.com/profile/02747099358519893177noreply@blogger.comtag:blogger.com,1999:blog-35364652.post-2688150199862670002020-08-14T13:59:41.007-04:002020-08-14T13:59:41.007-04:00Hi Ernie.
Interesting post. Would this work on sub...Hi Ernie.<br />Interesting post. Would this work on sub-daily frequency features? I have a strategy that holds asset for a few days, but checks conditions intraday. <br /><br />Also - would having Binary features be a problem? Ankit Garghttps://www.blogger.com/profile/08202373781470135289noreply@blogger.comtag:blogger.com,1999:blog-35364652.post-90173328181414365602020-08-12T07:38:53.342-04:002020-08-12T07:38:53.342-04:00Hi Alex,
Indeed, the features you use for metalabe...Hi Alex,<br />Indeed, the features you use for metalabeling should be different (not necessarily uncorrelated or orthogonal) from the indicators you already use in your base strategy. Otherwise, metalabeling can't add too much value to your current strategy. The idea is to scour the world of predictors that can tell you whether your strategy is in a good "regime" to trade or not, predictors that are not already part of your strategy.<br /><br />Hope this makes sense?<br />Ernie<br />Ernie Chanhttps://www.blogger.com/profile/02747099358519893177noreply@blogger.comtag:blogger.com,1999:blog-35364652.post-63629063823412383452020-08-12T04:14:35.917-04:002020-08-12T04:14:35.917-04:00Hi Ernie,
One follow-up question on how to use th...Hi Ernie,<br /><br />One follow-up question on how to use the service. I have a bunch of features that I created from which I create a strategy. So in my case, the strategy is a function of all the indicators I have (i.e: the average of all the indicators). As I understood the service, the indicators are the features and the strategy output the labels.<br /><br />When I read the post, I was under the impression that the strategy must be somewhat uncorrelated to the features. In other words if f(x) is the strategy, then x should not include the indicators used as the features of the model prediction. I assume that otherwise we end up with a conditional probability.<br /><br />Have I gotten this wrong?<br /><br />ThanksAlexnoreply@blogger.comtag:blogger.com,1999:blog-35364652.post-75928414399210400832020-08-11T11:58:04.369-04:002020-08-11T11:58:04.369-04:00Mark,
Yes, if your filters are useful for predicti...Mark,<br />Yes, if your filters are useful for predicting whether your signals will be profitable, the predictive accuracy and AUC score should be higher than 50%.<br /><br />Please feel free to email me at ernest@predictnow.ai (with your predictnow.ai username), and we will take a look at your files to see if there is anything amiss.<br /><br />Thanks,<br />ErnieErnie Chanhttps://www.blogger.com/profile/02747099358519893177noreply@blogger.comtag:blogger.com,1999:blog-35364652.post-31364338430987443262020-08-11T11:28:49.101-04:002020-08-11T11:28:49.101-04:00Shouldn't predicted direction accuracy > ba...Shouldn't predicted direction accuracy > baseline win rate?<br /><br />baseline 54-57% win rate -> 66-67% filtered via backtest win rate -> PredictNow.ai 48-49% win rate (predicted direction accuracy vs backtest actual p/l)<br /><br />Ernie, I spent few hours playing with the PredictNow.ai tool and listened to your June presentation on it. I either am not doing it right, or it's not adding value to models I thoroughly backtested and am live trading. The 2 main models I tried are based on capturing the overnight return in certain equities, buying the close and selling the next open. In the baseline, taking all signals, the win rate would be 54-57%. By adding 10-30 filters and taking only the most productive 1/3rd of the signals, the win rate goes up to around 66-67%. Enter PredictNow.ai, I took my daily data with the variables used in my filters, and the training data results are all below 50% accuracy (ACCURACY SCORE FOR CV = 0.49; AUC SCORE FOR CV= 0.48). Shouldn't predicted direction accuracy > baseline win rate?<br /><br />Marknoreply@blogger.comtag:blogger.com,1999:blog-35364652.post-51953177561862297412020-08-10T09:30:12.133-04:002020-08-10T09:30:12.133-04:00Hi Ernie -
You are doing some interesting things!...Hi Ernie -<br /><br />You are doing some interesting things!Kevin Daveyhttps://www.blogger.com/profile/04236816654726712466noreply@blogger.comtag:blogger.com,1999:blog-35364652.post-84286614791316616002020-08-09T16:56:34.175-04:002020-08-09T16:56:34.175-04:00Hi John,
Yes, I heard that HMM already performed w...Hi John,<br />Yes, I heard that HMM already performed well in predicting tail regimes.<br />ErnieErnie Chanhttps://www.blogger.com/profile/02747099358519893177noreply@blogger.comtag:blogger.com,1999:blog-35364652.post-37119195863831741502020-08-09T10:59:15.563-04:002020-08-09T10:59:15.563-04:00Hi ernie,
Yes, i just read your tail pape...Hi ernie,<br /> Yes, i just read your tail paper . That provides a ton of context.<br /> In "machine trading " you wrote about hidden markov models to tell whether we are in bull or bear market . Have you seen the model giving trouble pointers during march collapse? Or do you recommend for me to test that strategy too on s&p ? <br /> Basically does HMM have the potential to be used as a tail risk strategy?<br /> Thanks again for all your teachings . It has been so much valuable in my trading journey!<br />Best,<br />John.Johnhttps://www.blogger.com/profile/08323598876861557631noreply@blogger.comtag:blogger.com,1999:blog-35364652.post-88384699639370556832020-08-09T10:53:17.013-04:002020-08-09T10:53:17.013-04:00Hi Alex,
I recommend re-training whenever your dat...Hi Alex,<br />I recommend re-training whenever your data changes by 5% or more.<br />ErnieErnie Chanhttps://www.blogger.com/profile/02747099358519893177noreply@blogger.com