tag:blogger.com,1999:blog-261139923818144971.post7894682886656168506..comments2024-03-27T07:58:49.946+00:00Comments on This Blog is Systematic: Does it make sense to change your trading behaviour in different periods of volatility?Rob Carverhttp://www.blogger.com/profile/10175885372013572770noreply@blogger.comBlogger38125tag:blogger.com,1999:blog-261139923818144971.post-5273198485344000882022-05-02T18:09:44.166+01:002022-05-02T18:09:44.166+01:00Very hard to measure volumes for futures because o...Very hard to measure volumes for futures because of rolls. I've never found any systematic effects of volume on price.Rob Carverhttps://www.blogger.com/profile/10175885372013572770noreply@blogger.comtag:blogger.com,1999:blog-261139923818144971.post-59393648349559733252022-05-02T13:13:15.358+01:002022-05-02T13:13:15.358+01:00Hi Rob,
Thanks for the post. Slightly unrelated q...Hi Rob,<br />Thanks for the post. Slightly unrelated question – Do you have any views on how trading volumes affect trend following? <br />How would one go about investigating this question in the same format you have done above for volatility? E.g. how do you recommend in measuring trading volumes<br />Anonymous https://www.blogger.com/profile/14019067484620273077noreply@blogger.comtag:blogger.com,1999:blog-261139923818144971.post-3939106140811103782021-05-21T12:17:18.440+01:002021-05-21T12:17:18.440+01:00Yes except you'd probably want to Z-score rath...Yes except you'd probably want to Z-score rather than just demeaning.Rob Carverhttps://www.blogger.com/profile/10175885372013572770noreply@blogger.comtag:blogger.com,1999:blog-261139923818144971.post-18792471633138008312021-05-21T12:12:10.843+01:002021-05-21T12:12:10.843+01:00Thanks Rob.
Just to understand this a bit better ...Thanks Rob.<br /><br />Just to understand this a bit better in the meantime. Assuming I want to implement a Value signal using country ETFs. <br /><br />A) UK equity earnings yield = 8%<br />B) Australia equity earnings yield = 4%<br />C) MSCI World ewnrjtjd yield = 6%<br /><br />“Relative Value” Raw Forecast UK Eq = -(8%-4%)-6%= +2%<br /><br />Relative value raw forecast Aussie Eq= [-(4%-8%)]-6%= -2%<br /><br />Is my interpretation correct?<br /><br />Thanks,<br />RUnknownhttps://www.blogger.com/profile/05821050393149046409noreply@blogger.comtag:blogger.com,1999:blog-261139923818144971.post-37072411685510798502021-05-21T11:59:51.226+01:002021-05-21T11:59:51.226+01:00For eg any pair trading you create a synthetic ins...For eg any pair trading you create a synthetic instrument that embodies the spread, then your forecast would be something like -(price of spread - fair value), and then you translate the synthetic instrument back to actual positions. I'll probably blog about this at some point.Rob Carverhttps://www.blogger.com/profile/10175885372013572770noreply@blogger.comtag:blogger.com,1999:blog-261139923818144971.post-7828927157288053962021-05-21T11:50:27.113+01:002021-05-21T11:50:27.113+01:00Hi Rob - I hope you are well.
Would you mind expl...Hi Rob - I hope you are well.<br /><br />Would you mind explaining how your system and the calculation of forecasts can be adapted if we were to try and implement a relative value strategy / stat arb strategy or signal? <br /><br />Thanks in advance for your help.Unknownhttps://www.blogger.com/profile/05821050393149046409noreply@blogger.comtag:blogger.com,1999:blog-261139923818144971.post-27280487852063889962021-05-17T10:31:58.295+01:002021-05-17T10:31:58.295+01:00Thanks Rob.
I thought about it but it felt too si...Thanks Rob.<br /><br />I thought about it but it felt too simple to me in the first instance, hence I wanted to get your opinion. <br /><br />Thanks Unknownhttps://www.blogger.com/profile/05821050393149046409noreply@blogger.comtag:blogger.com,1999:blog-261139923818144971.post-23353538443741129122021-05-17T10:25:59.175+01:002021-05-17T10:25:59.175+01:00Well you just calculate the forecasts, and then su...Well you just calculate the forecasts, and then subtract the average forecast for each sector... right?Rob Carverhttps://www.blogger.com/profile/10175885372013572770noreply@blogger.comtag:blogger.com,1999:blog-261139923818144971.post-17272663166787558432021-05-17T10:10:36.379+01:002021-05-17T10:10:36.379+01:00Yes. My benchmark could be the FTSE 100 and I have...Yes. My benchmark could be the FTSE 100 and I have the ability to buy the underlying FTSE 100 GICS Sector ETFs, 11 of them - energy, consumer staples, industrials etc. We are in a market environment where most of the sectors have strong up trends but some of them have weaker trends. <br /><br />How can we compute the forecasts in such a way that the sum of the deviations of each sector ETF from BM is equal to zero? <br /><br />The mandate is like beating the FTSE 100 with a TAA approach, where at each point in time we will be long some sectors and short some others but the sum of all the deviations is zero. <br /><br />What’s the best practice to achieve that?<br /><br />Thanks Unknownhttps://www.blogger.com/profile/05821050393149046409noreply@blogger.comtag:blogger.com,1999:blog-261139923818144971.post-50701433016573123942021-05-17T09:53:32.136+01:002021-05-17T09:53:32.136+01:00So I assume you are trying to put together a long ...So I assume you are trying to put together a long only type portfolio? And do you want to buy a selection with equal weights (eg 'buy 10 out of the FTSE 100') or in theory buy everything with different weights?Rob Carverhttps://www.blogger.com/profile/10175885372013572770noreply@blogger.comtag:blogger.com,1999:blog-261139923818144971.post-33991072174104004582021-05-15T11:31:20.328+01:002021-05-15T11:31:20.328+01:00Thanks Rob.
If I was going to build a trend follo...Thanks Rob.<br /><br />If I was going to build a trend following model like yours but using a set of country etfs - as many as there are in the chosen equity bm - with the idea of beating the MSCI ACWI, then how can I account for the fact that forecasts might signal a buy trade across the board for all of them, if all the equity markets are trending in the same direction? <br /><br />Do I have to calculate an average forecast across them and then subtract such an average from each forecast? <br /><br />Or shall I subtract the minimum value between forecasts from each of them? <br /><br />I hope I have been clear where I am coming from. Your trading system is absolute oriented, whilst I am trying to beat a defined benchmark and I wonder how I can adapt your system to that.<br /><br />It would be great if you could share what the best practices are in such situation. <br /><br />Thanks in advance. Unknownhttps://www.blogger.com/profile/05821050393149046409noreply@blogger.comtag:blogger.com,1999:blog-261139923818144971.post-87087475451133478042021-05-10T20:37:20.822+01:002021-05-10T20:37:20.822+01:00I'm not sure. I've never seen a definition...I'm not sure. I've never seen a definition of what 'global macro' investing is. I guess it's investing at a global level, so equity indices, currencies... things you buy with futures? And 'macro', does that mean only using macroeconomic indicators? Or a broader set of signals? I guess one could use a trend following system as part of that.<br /><br />Answer to your second question is yes, absolutely. It's less relevant for relative value / stat arb / ... however, but can be made to work with some modifications.Rob Carverhttps://www.blogger.com/profile/10175885372013572770noreply@blogger.comtag:blogger.com,1999:blog-261139923818144971.post-19748964783644506842021-05-10T20:25:03.084+01:002021-05-10T20:25:03.084+01:00Hi Rob - how would you differentiate trend followi...Hi Rob - how would you differentiate trend following from global macro investing? <br /><br />Also, could one apply your trading system framework to other types of investing - I.e. fundamental investing - or do you think raw forecasts, forecast scalars, etc will have to be calculated differently? <br /><br />Thanks,<br />RUnknownhttps://www.blogger.com/profile/05821050393149046409noreply@blogger.comtag:blogger.com,1999:blog-261139923818144971.post-36592140460339421392021-04-19T09:07:19.866+01:002021-04-19T09:07:19.866+01:001) Yes, although in practice I do it in a backward...1) Yes, although in practice I do it in a backward looking way in the backtest.<br /><br />2) What I'm doing is actually pretty similar to a time series Z score. The main difference is that in a Z score you would subtract the mean. I don't want to do that; I want to preserve any long/short bias in the forecast. For example, consider a slow momentum on an asset that has been going up through most of history. The raw forecast will mostly be long that asset, and I want to preserve that 'longness'. The other minor difference is that a Z score uses the standard deviation and I use the mean absolute deviation.<br /><br />3) The only reputable course I am aware of is this https://www.sbs.ox.ac.uk/programmes/executive-education/online-programmes/oxford-algorithmic-trading-programme <br /><br />Of course I haven't done the course, but the list of contributors is pretty top notch.Rob Carverhttps://www.blogger.com/profile/10175885372013572770noreply@blogger.comtag:blogger.com,1999:blog-261139923818144971.post-86854735461152919312021-04-17T14:59:59.417+01:002021-04-17T14:59:59.417+01:00Hi Rob - first of all, it was great to see you on ...Hi Rob - first of all, it was great to see you on YouTube yesterday and discuss the systematic topics. Thanks for keeping educating us. <br /><br />A couple of questions on the forecast scalars:<br /><br />1) in your book and supporting spreadsheets, you advise to use a “fixed” scalar for a given trading rule variation. May I know how you fit these? Let’s suppose I have 10 years of backtest with related raw forecasts - do you use the entire 10 years to come up with your “fixed forecast scalar”, which will then be applied to live trading system? <br /><br />2) why did you not use a Z score methodology to transform raw forecasts into normalised forecasts? Is the Z-score methodology an inferior one to the one proposed by you?<br /><br />3) if I wanted to get training and a certification in systematic trading, what steps would you advice?<br /><br />Thanks in advance for your replies.<br /><br />RiccardoUnknownhttps://www.blogger.com/profile/05821050393149046409noreply@blogger.comtag:blogger.com,1999:blog-261139923818144971.post-14237805112973664352021-04-12T13:45:51.679+01:002021-04-12T13:45:51.679+01:00Yes on reflection, this would probably make more s...Yes on reflection, this would probably make more sense than the version I described.Rob Carverhttps://www.blogger.com/profile/10175885372013572770noreply@blogger.comtag:blogger.com,1999:blog-261139923818144971.post-73773128666856546752021-04-12T10:49:54.873+01:002021-04-12T10:49:54.873+01:00Apologies for not being clear. For “adjusted forec...Apologies for not being clear. For “adjusted forecast”, I mean the “raw forecast” adjusted by “forecast scalars”. <br />So, what if we applied “L” to this “adjusted forecast”? Would that be acceptable?<br /><br />Thanks!Unknownhttps://www.blogger.com/profile/05821050393149046409noreply@blogger.comtag:blogger.com,1999:blog-261139923818144971.post-23647691502518319692021-04-12T10:46:43.002+01:002021-04-12T10:46:43.002+01:00What's the 'adjusted' forecast??What's the 'adjusted' forecast??Rob Carverhttps://www.blogger.com/profile/10175885372013572770noreply@blogger.comtag:blogger.com,1999:blog-261139923818144971.post-20775420488601987052021-04-12T10:40:15.542+01:002021-04-12T10:40:15.542+01:00Thanks for your reply, Robert.
Follow-up question...Thanks for your reply, Robert.<br /><br />Follow-up question: what if we applied “L” to the “adjusted forecast” instead? Would that be acceptable ?<br /><br />Thanks, RicUnknownhttps://www.blogger.com/profile/05821050393149046409noreply@blogger.comtag:blogger.com,1999:blog-261139923818144971.post-12583518571682092232021-04-12T08:50:49.675+01:002021-04-12T08:50:49.675+01:00You've been clear, but you are missing the fac...You've been clear, but you are missing the fact that the forecast scalars are effectively fixed.<br /><br />That means they won't 'rein in' risk as you describe, since we're going to multiply the modified forecast by the same scalar regardless of environment.<br /><br />Now for the disclaimer: they're not exactly fixed, since they are estimated on a rolling out of sample basis. That means if we start in say an (ex-post) low vol environment then the the forecast scalars will be too low (since the raw forecast is higher, and the scalar doesn't need to be as large); but after a few years it should be sorted out.Rob Carverhttps://www.blogger.com/profile/10175885372013572770noreply@blogger.comtag:blogger.com,1999:blog-261139923818144971.post-22995569496791454222021-04-10T12:20:49.398+01:002021-04-10T12:20:49.398+01:00Hi Rob - I would like to ask you a clarification o...Hi Rob - I would like to ask you a clarification on the following statement:” I use the raw forecast here. I do this because there is no guarantee that the above will result in the forecast retaining the correct scaling. So if I then estimate forecast scalars using these transformed forecasts, I will end up with something that has the right scaling”.<br /><br />If we apply “L” to the “raw forecast” and then we apply the “forecast scalars” to the raw forecast adjusted by L, then are we not diluting / losing the overall signal coming from L? I mean, the overall signal might well say - hey, it’s a high vol environment, let’s rein in risk; but then our forecast scalars will say, No, we need to adjust the position upward to make it more like an average bet, whisk we don’t want to take an average bet in the forecast adjusted for L is low.<br /><br />I hope I have been clear. If not let me know. <br /><br />Thanks,<br />RicUnknownhttps://www.blogger.com/profile/05821050393149046409noreply@blogger.comtag:blogger.com,1999:blog-261139923818144971.post-11457563938661847182021-04-03T12:56:21.897+01:002021-04-03T12:56:21.897+01:00Thank you for the great post.Thank you for the great post.Anonymous https://www.blogger.com/profile/14019067484620273077noreply@blogger.comtag:blogger.com,1999:blog-261139923818144971.post-64241608501791557282021-03-11T21:44:36.069+00:002021-03-11T21:44:36.069+00:00Thanks for clarifying, Rob. I was thinking whether...Thanks for clarifying, Rob. I was thinking whether it could be some sort of weird Z-score-like measure. Fortunately, it is just a pasting error.Rafael Ladeirahttps://www.blogger.com/profile/04888748113522914330noreply@blogger.comtag:blogger.com,1999:blog-261139923818144971.post-85272276852609397502021-03-11T19:11:38.165+00:002021-03-11T19:11:38.165+00:00Wow. I expected trading costs to be a more signifi...Wow. I expected trading costs to be a more significant factor than they are. Not necessarily in the general decline in performance but rather in the relative poor performance of faster trading rules. But costs don't seem to be significant here either (if my math is right). <br /><br />Either way, it seems likes the best thing to do is to scale back your positions during periods of high vol. Of course, those who dynamically vol target/scale are already doing this.<br /><br />Thanks Rob for doing the research, and the extra research too.Michael Newtonhttps://www.blogger.com/profile/15187744343817723912noreply@blogger.comtag:blogger.com,1999:blog-261139923818144971.post-10886788749785501592021-03-11T10:20:10.910+00:002021-03-11T10:20:10.910+00:00Updated the post to reflect this.Updated the post to reflect this.Rob Carverhttps://www.blogger.com/profile/10175885372013572770noreply@blogger.com