OmniVest User Forums
User-Created Portfolios
OV Account Analysis Tool
21 replies, 2043 viewings
|
OmniVest User Forums
User-Created Portfolios
OV Account Analysis Tool
|
Last Activity 7/22/2015 8:11 AM 21 replies, 2043 viewings |
|
|
|
| Printer friendly version |
Steve2 Elite    Posts: 750 Joined: 10/11/2012 Location: Annapolis, MD  User Profile |
The "Positions" worksheet contains instructions for using the workbook and is where your positions data is pasted. The "Strat Analysis" worksheet breaks down realized performance by calendar quarter for the overall account as well as for each individual strategy. We thought the strategy break out would be useful for tailoring strategy composition but we found that without visibility into how OV constructs its sorted list of trade candidates each day, strategy refinement is really more like playing whack-a-mole. It's pretty fascinating to see how the strategies interact. Hopefully in the future, Nirvana will provide more visibility and perhaps more user control over how their sorting algorithm works. If so, then we expect the strategy break out will be quite useful. Also on this worksheet are a number of account statistics designed to capture account volatility (e.g., counts of positions by P/L ranges, largest consecutive winning/losing streaks, P/L % volatility, etc) The "Sym Analysis" worksheet breaks out each unique symbol traded along with how frequently it was traded and it's contribution to P/L. This enables you to assess whether or not a significant percentage of your P/L is concentrated in a relatively small number of symbols which we view as risky. We also think this break out may be useful for refining focus lists once OV releases the user-defined focus list capability. The "Key Metrics" worksheet summarizes our optimization approach and pulls some key metrics from other parts of the workbook that are used for performing this optimization. Caveats: The workbook uses macros so you have to enable macros. It was developed in Excel 2007. I don't know if it will work in earlier (or for that matter later) versions of Excel. You need a big-boy PC to run it. For example, it takes about 2.5 minutes to calculate the stats for a 14 year simulation with 26,000 positions using a 2.5 GHz Quad-Core PC. Other than that, it's pretty simple to use. Just paste in your data and click the "Prepare Data" button and out come the stats. If you decide to play with this, please let us know: 1. If you find any errors 2. Have suggestions for additional useful statistics 3. Have any comments on the optimization methodology we're using (we have very thick skin :)) Enjoy! Steve and Ian [Update] The attachment has been removed since something in the upload/download process is corrupting the file. Please you this dropbox link to download a clean version of the file: https://www.dropbox.com/s/wpld9rnfmsxtduh/Ov%20Sim%20Analysis%20Tool%20v1.12.xlsm [Edited by Steve2 on 1/25/2014 8:46 AM] | ||
| Mark Holstius Elite Posts: 744 Joined: 10/11/2012 Location: Sleepy Hollow, IL User Profile |
I just downloaded your spreadsheet and plan on trying it out this weekend - it looks like it'll have a lot of useful information & I appreciate the work involved in developing it. Very nice of you to post it. Mark | ||
| Fred Gordon Legend Posts: 481 Joined: 10/11/2012 Location: Fayetteville, Ga User Profile |
Fred I guess that answers that; just downloaded it but my converter will not save it to my Excel ver. Will try again when new computer arrives. [Edited by Fred Gordon on 1/24/2014 9:53 AM] | ||
| George Member Posts: 24 Joined: 9/29/2013 Location: Independence, Minnesota User Profile |
| ||
| Steve Mayo Legend Posts: 414 Joined: 10/11/2012 Location: Austin, TX User Profile |
Steve | ||
| Steve2 Elite Posts: 750 Joined: 10/11/2012 Location: Annapolis, MD User Profile |
https://www.dropbox.com/s/wpld9rnfmsxtduh/Ov%20Sim%20Analysis%20Tool%20v1.12.xlsm Also, I searched for issues reading Excel 2007 files from Excel 2013 and there are some. One suggested workaround was to start Excel 2013 first and then open the file from within Excel rather than just double-clicking on the file to launch Excel. Steve | ||
| John W Elite Posts: 654 Joined: 10/11/2012 Location: Sydney, NSW, Australia User Profile |
THANK YOU for your and Ian's generosity! WELL DONE! | ||
| Steve Mayo Legend Posts: 414 Joined: 10/11/2012 Location: Austin, TX User Profile |
Thank you for sharing your spreadsheet. It does an excellent job of showing how the component strategies and symbols contribute to the overall portfolio performance. I like that it highlights how just a few stocks generate the bulk of the return in most portfolios. In the portfolio I tested, for example, it was interesting to note that 50% of the return came from only 28 stocks out of the 500+ that were traded! Mark Holstius and I have been working on portfolio optimization for almost a year now, using an approach that involves automated testing of thousands of strategy combinations. Others in the forum have reported using the recent short-term best-performers in sort of a momentum approach. Keith McIntyre has been promoting sequential strategy assembly. There are lots of interesting approaches. The difficulty I find with all these various approaches, however, is the question of predictability. Can we be confident that our candidate portfolio will again identify those 28 or so high-flyers next year? I note that the standard deviations of the monthly returns are multiples of the means on most, if not all, strats/ports I have studied. Taking ARM4 Margin as an example of a good portfolio, there is a 5% probability of a 16% (2 Sigma) drop each month for a portfolio that averages about 5% return/month. [This assumes a normal distribution whereas the distribution graph shows the port actually has fatter tails.] In short, there's a lot of variability in these portfolios. Of course, that's a good thing on the upside. :-) To get back on point, reviewing your spreadsheet, it appears your approach is to identify strategies for replacement within a candidate portfolio, seeking to maximize consistency foremost. For comparison purposes with other "optimizers", would you be willing to share some metrics from your best-performing optimized portfolio so we can compare? I would also be interested to learn how well your live-trading results match your optimized model. Meanwhile, I will share some of our candidate portfolios and risk measures; we haven’t yet traded these live yet. The attached graphs show some of the metrics used to compare 3 recent candidates against Nirvana’s ARM4 Margin portfolio (our benchmark) as well as the overall market. Our best candidate in this series (2013-1-1 v2) has a 6-year CAGR of 90%, MDD of 32% for a Calmar of 2.8. The Modigliani and Sortino ratios show it is the better candidate of this group on a risk-adjusted, benchmark-relative basis. However, the non-parametric box plot and the overlapping confidence intervals show that, statistically-speaking, the difference in mean monthly returns between these 4 portfolios are not significantly different (p=0.6 by ANOVA). Any material different result one might get trading any of these would most likely be random chance. Statistics aside, I think you are on the right track with making consistency your primary optimization metric. Optimizing for maximum return alone, in my opinion at least, is likely to result in a portfolio that will not repeat its luck going forward. [Edited by Steve Mayo on 1/27/2014 4:06 PM]  Attached file : OV - Detailed Analysis of November Candidates.docx (232KB - 367 downloads) | ||
| Steve2 Elite Posts: 750 Joined: 10/11/2012 Location: Annapolis, MD User Profile |
Thanks for sharing your results. I agree that the key question for any of these approaches is how similar future returns will be to past performance. Our focus has been to create portfolios that provide consistently positive past performance and not necessarily the highest returns. Only time will tell whether or not this will work :-). We have not yet begun live trading the results of this work. Attached are the results of a 6-year simulation with the portfolio we plan to go live with in February. Steve [Edited by Steve2 on 1/27/2014 7:05 PM] Attached file : Quarterly Results.jpg (234KB - 341 downloads) Attached file : Portfolio Stats.jpg (150KB - 323 downloads) | ||
| Flying Dutchman Member Posts: 24 Joined: 3/17/2014 Location: The Netherlands User Profile |
Thank you so much for setting up this Excel sheet, this really helps a lot! FYI - also works fine in Office 2010. One suggestion, if I may: would it be possible to add a profit-factor to the metrics as well? And , perhaps, some distribution chart of P/L-percentages (to see whether it would more or less fit a Bell curve) ? Thanks again!! | ||
| Steve Mayo Legend Posts: 414 Joined: 10/11/2012 Location: Austin, TX User Profile |
I did some work with distribution curves and Monte Carlo simulations on the OmniVesting.com website. Look in the Tools section for "Portfolio Analyzer". I'm working on an updated version to make it easier to use. Meanwhile, let me know if you need me to walk you through it. :-) Steve | ||
| Steve2 Elite Posts: 750 Joined: 10/11/2012 Location: Annapolis, MD User Profile |
The spreadsheet tool was developed over a year ago and I'm afraid we're no longer making updates (except to correct errors). I would caution you not to read too much into the quarterly P/L results of a single simulation. It is useful for doing initial account optimizations but in order to get reasonable confidence in P/L distributions, you need to do a bunch (hundreds) of quarterly simulations using random starting dates within your simulation range. The issue is that most portfolios will generate far more trades each day than you have account equity to take. This means that each simulation (using a different starting date) will have a unique trade history that typically varies throughout the simulation period not just at the beginning and end due to a change in starting/ending date. While I haven't tried Steve Mayo's tool yet, it sounds like his tool is providing this type of analysis so it would definitely be worth checking out. Steve | ||
| Steve Mayo Legend Posts: 414 Joined: 10/11/2012 Location: Austin, TX User Profile |
Basically, as Steve2 points out, the backtest is only ONE possible outcome, with nothing to tell you how likely THAT particular outcome might be. You really can't make a statistically-sound judgement based on that single equity curve, or those single-number metrics like CAR and MDD without knowing the probability range, specifically mean and standard deviation, or median and interquartile range. Even with this tool, I would still caution you to do what Steve2 suggests as well. Rerun your simulation with differing start dates, differing "Sort By's, and different symbol lists to get a better feel for the likely performance. Also, it's best to allocate "blocks" of funds to each strategy so that one strategy doesn't compete so much with another, and then keep the allocation/trade low so that your strategies only rarely ever hit the equity ceiling. | ||
| Steve2 Elite Posts: 750 Joined: 10/11/2012 Location: Annapolis, MD User Profile |
I'm not convinced that the strategy outlined in your last paragraph is the best approach. I tend to try and optimize for consistent performance per period, high %Win, and high %Invested (which typically means high TPM) and don't worry much about draw downs. While I think your approach would improve consistency, I'm not sure it would necessarily deliver very high returns. But then I'm sure you and Mark have done analyses that show otherwise... right? :-) Steve [Edited by Steve2 on 4/21/2015 7:50 PM] | ||
| Steve Mayo Legend Posts: 414 Joined: 10/11/2012 Location: Austin, TX User Profile |
In OV, the only (readily-available) measure we have of that return:risk relationship is Calmar (CAR/MDD) (return above portfolio risk only). The problem, as you described well, is that it is hard to trust those metrics when they can change dramatically with only a minor settings tweak or a slight change in start/end date. Your approach of looking at consistency of periodic segments is certainly better, but it still has the problem where portfolio return:risk changes dramatically once you start hitting the equity ceiling. I'm essentially doing the same thing, but I use much smaller "periods" and I don't assume they will occur in the same sequence or even that all of them will occur every time. It doesn't entirely fix the equity ceiling problem per se, but it does give a much better estimate of what the is true overall portfolio performance. Once I get version 2 finished, I'll try to better explain the concept. :-) | ||
| Flying Dutchman Member Posts: 24 Joined: 3/17/2014 Location: The Netherlands User Profile |
I appreciate your answer, but on visiting the Omnivesting.com website, I just get an error: THIS REQUEST IS BLOCKED BY ADMIN TOOLS. PLEASE CHANGE THIS MESSAGE IN THE COMPONENT'S OPTIONS. Seems the site (is it yours?) is down for some reason. | ||
| Steve Mayo Legend Posts: 414 Joined: 10/11/2012 Location: Austin, TX User Profile |
If you want to send me an export for your equity data (from the portfolio page, preferably), I'll send you back the report. You can send via the private posting if you don't want to share it publicly. If you do that, also send me your IP number so I can chance down why you got blocked. Thanks, Steve | ||
| Flying Dutchman Member Posts: 24 Joined: 3/17/2014 Location: The Netherlands User Profile |
| ||
| Steve Mayo Legend Posts: 414 Joined: 10/11/2012 Location: Austin, TX User Profile |
| ||
| Fred Gordon Legend Posts: 481 Joined: 10/11/2012 Location: Fayetteville, Ga User Profile |
Do you think it is safe to click "OK"? Thanks, Fred Gordon | ||
| Flying Dutchman Member Posts: 24 Joined: 3/17/2014 Location: The Netherlands User Profile |
Well - I'm not a financial analist, that would be way too much credit. Instead, I'm into fiscal advisory services and accounting. Yeah - that really dull stuff :-) As to the portfolio: I've actually not taken any live trades with OV. Even after a year, I am still trying to get my head around a few confusing concepts. The work that Ed's team has been doing is amazing, mindblowing but also very confusing if you don't spend hours and hours on learning what to do with the tools, and what not to do / what not to expect. (For example : it took me quite a while to realize that having selected an account from the accounts page, and then going into "Tools" and messing around a bit with some tools, "adding" something, I would not even realize that I'd be adding that stuff to THAT particular account that I "was on"... (without any account name mentioned on a page..) It took me some darn time to figure out the entire OV eco-system, I can tell you that... And, in general, I tend not to regard myself to be all that stupid... ;-) Also, allow me to say: Steve, thanks for setting up your omnivesting.com site. I am without doubt that I can learn much more from your site on a proper usage of OV. THANK YOU, STEVE ! As a small "hail to thee" I have just ordered your "NASDAQ-100 HISTORIC LIST IN OMNIVEST FORMAT". I had not realized before, that we could end up SO far off, simply on a static Nasdaq100 list... Ouch. Anyway, so I can't tell you from personal OV live trading experience. And I am still hesitant to put it to real life work. You see, sure I can create a USD 14 Billion 2000-2015 equity outcome, 94 trades a month, a CAR of 116% and a CALMAR of 3.2 , 67% profitable. But I just don't trust such figures. And I shouldn't either :) So in the end it is all about this: can I be using OV to make real trades in a proper way... Or just stick to more simple systems, good old OmniTrader ;-) and be managing trades. Since I tend to believe that spending my hours on managing (perhaps not the best) trades might get me further than spending those hours and hourse on endless equity curves that do not represent real life in the end. So I guess this is my way of saying: I'm still not convinced that I can get OV work for me as the ultimate trading vehicle ;-) What I have also found: making the slightest adjustments to the Account Settings (like capping the trades to max 2% of daily volume) will have considerable impact on any outcome. Anyway , just a few thoughts, and I now shall have a great time reading the material on your website! Thanks again. | ||
| Steve Mayo Legend Posts: 414 Joined: 10/11/2012 Location: Austin, TX User Profile |
But, to Nirvana's outstanding credit, this is cutting-edge never-done-before stuff! They are pioneers blazing a trail into unchartered territory. There's just no way to avoid a winding trail when you are doing that. :-) And they are working hard on trying to make it easier and more intuitive. They've created OmniPortfolios as a first step and are, as well, hard at work on a new tool that will automate much of the "building" process. It just takes time and money, and its oh so much harder when working on a "live" system. I'm part of a team that meets every week just to talk about such stuff and we often end up going round-and-round trying to decide which direction to go: less complicated or more functional -- its hard to do both at the same time with no prior examples to follow! :-) |
|
|
|
|
Administrator
Forum Moderator |
Registered User
Unregistered User |
 E-Mail this thread to a friend |
Toggle e-mail notification |
