OmniTrader Upgrade Forums
ATM
A Confidence Boost
26 replies, 3581 viewings
|
OmniTrader Upgrade Forums
ATM
A Confidence Boost
|
Last Activity 4/21/2021 9:48 AM 26 replies, 3581 viewings |
|
|
| Printer friendly version |
mholstius Veteran   Posts: 175 Joined: 1/13/2017  User Profile |
A trader may have wonderful historical results to share, but once it’s uploaded we all get to witness whether the concept works at the right edge (or not). After it’s published, you can’t cherry pick data to display on the forum - it’s out there for everyone to see and critique. I’ve developed extensive spreadsheets to extract data & statistics to test systems and concepts as much as possible before posting them. But no matter how stable, consistent, and robust it may be in the past, there’s always the question of whether a portfolio will continue to trade consistently at the right edge. We discuss it a lot here: robustness vs performance, curve fitting vs ranking. The confirmation is in the trading after the rules and systems are published and open to peer review. The discussion of “out of sample” results becomes moot: everything after publishing is obviously out of sample. Back on 6/21/18, I posted the results from combining 2 Elite portfolios in OV: Follow The Money HD and ATM Macro & Micro here; https://www.omnitrader.com/currentclients/omnivestforum/thread-view.asp?threadid=7880&posts=8 Combining the two portfolios looked particularly promising, and I went into great detail about how they “compliment” each other - taking trades at different times & utilizing account equity at a high level. I’ve always thought that a system with numerous trades under varying market conditions, that doesn’t attempt to avoid the “black swans” in the past, will probably recover from the inevitable DDs and challenges in the future. With extensive testing, I feel that the risk of ruin with a system developed this way is negligible - so I treat the portfolios I create using Nirvana’s excellent tools like mutual funds. I let them trade autonomously, without interference or tweaks, in order to give them time to “prove themselves”. That’s easy to say when you’re trading your own account, but I feel a heightened (uncomfortable?) sense of responsibility after I’ve published something and folks are actually subscribing to trade the system. It’s great when things are going up, but especially disconcerting when riding out one of those inevitable drawdown periods that are so easy to overlook in a long term logarithmic chart - and knowing that other people are enduring the DD with you (or because of you). The 2 months from August 28th to Oct 29th ended up being one of those “reality” periods for this system. It’s just part of the game, but a lot easier to contemplate than endure. Below is a snag of the 1 year OV performance using the identical settings, portfolios, etc., that I posted in June - along with the SPY for the same period; Everything was great from June to August (A to B), but it was difficult for the 2 months starting on Aug 28th, with a MTM DD of 21.7% on Oct 29th. I posted the following comment back on 6/21 about my hope that the ATM portion of the portfolio might be a good hedge in the future… and it’s been proven to be true in October. There’ve been 19 trades opened and closed since 10/29. 10 of them were ATM shorts, with only 1 loss among the 19 (a -0.34% short); (Note: it's a margin account, so those QTY% figures in OV are % of "base equity"... 24.6% is 12.3% of available margin equity) Given the drop in the market in October, a DD was expected - but I’m impressed by the way the portfolios (and particularly the ATM M&M) have adjusted to the changing market. SPY is at the same level it was on 6/21, but the system is up 14.6% since then. As I’ve said before, ATM is a game changer. I want to thank the traders that’ve subscribed via Elite, and it could be perceived as simply good PR for me to post that the portfolio’s performance is consistent with its history - but that’s not my primary reason for this post. It’s published, public, out of sample, confirming data for anyone questioning whether to trust the ability and performance of ATM. This system is based on numerous small RTM trades over long periods, and the results have been consistent. Given the numerous challenges and fluctuations this year, its performance should boost everyone’s confidence in the ability of ATM to adapt to changing markets. In June, you had to take my word for it that the last few years of data that I posted were out of sample - but there’s no denying that the results since then are. I want a little more time to improve it even more before I gift it to my grandchildren, but I’m especially pleased with how the results have improved since Nirvana developed ATM and gave us the ability to rank the trades. I plan to post more about my testing of a “walk Forward” method, and applying Concurrent Market States to ATM M&M, but I’ll be quite busy the next 2 months visiting children in California and Colorado and I wanted to get this out while I had the time. I’m definitely looking forward to getting OT 2019. I’m sure the advances included in it will make for an even more profitable New Year. I hope everyone has a wonderful holiday season… Mark [Edited by mholstius on 11/18/2018 7:51 PM]  Attached file : 00 OV Performance.png (322KB - 1849 downloads) Attached file : 01 June comment.png (93KB - 1783 downloads) Attached file : 02 Recent Trades.png (211KB - 1828 downloads) | ||
| sgbyrne Member Posts: 10 Joined: 1/2/2004 Location: Humble, Texas User Profile |
I really appreciate the research you have done in ATM and I am glad to see you are getting better results with combined strategies. I agree that you may have really good historical results but all too often strategies fail on the right edge, especially in a down-market. As far as I know Omnivest trades are 'MOO' on the day following the signal. My own research has shown that MOO trades are very unpredictable, I have put together examples of 'market' and 'MOO' trades using a single strategy with the same settings, with 'market' and 'MOO' stops. Ken Wilsdon, Nirvana Club Member, made a comment which gives one food for thought: 'I never use MOO orders. Only stop market, limit or stop limit orders. MOO orders do not give me confirmation that the security is moving in my direction at entry. I am at the mercy of the market makers on my MOO orders, and I am not in control. I could be entering much further away from my entry zone than I am comfortably doing, thus screwing up my risk/reward. Conversely, I would consider MOC orders, as this is the time when professional money is setting final value for a security and they rather than the market makers are in control.' Roll on ATM RT. Steve Byrne [Edited by sgbyrne on 11/23/2018 10:25 PM] Attached file : MarketOnOpen23Nov2018.jpg (570KB - 1722 downloads) Attached file : Market23Nov2018.jpg (540KB - 1654 downloads) | ||
| mholstius Veteran Posts: 175 Joined: 1/13/2017 User Profile |
I totally agree with both you & Ken in that MOO orders aren’t ideal, and it took me some time to put together the following test that might help with the discussion. I used a data set of 48,014 unique long trades from 1/1/2000 thru 8/14/18 generated by a variety of RTMs (CRT3, NSP-41, RTM7, T3-S3, VBX-3, XLS-19, XMFI, XULT) on 371 symbols. I used the MarketXLS extension for Excel to download the closing prices on the signal days of each trade (one day prior to the entry days and one day prior to the exit days). I calculated the return for each of the 4 possible combinations of MOO and MOC entries and exits and tabulated the results (thereby avoiding the effects of compounding returns); Column X calculates the difference in % sum of each combination vs the normal MOO entry and exit (#1). Columns Y-AB show how often a combination was either the best or worst outcome for a particular trade. The combination with the best return (#3: MOC Entry & MOO Exit) was the best choice for 32% of the trades and was the worst choice for 19% of them. The 2 pairs with MOC entries (#3 & 4) produced better results, but it’s interesting to note that MOO exits produced better results - when paired with either MOO or MOC entries. We can all remember times when we wished we’d gotten out of a trade the day before, but a MOO exit the next day was more beneficial in this set of 48,000 trades. Given all that, there are some serious practical problems with MOC orders in an autonomous system. With IB, a MOC order must be sent at least 15 minutes before the close, so you’d have to start the analysis at some point earlier based on the complexity of the system. Even a relatively simple system would probably need to be started 45 minutes before the close, and I really don’t know how much time would be required for something like my FTM portfolio containing 9 dynamic lists. Maybe someone who’s using MOC orders can comment on their experiences? We all know that a lot happens in the 30 minutes before close. I think it’s reasonable to expect that making a decision to enter a trade 30-45 minutes before the close would generate some false positives: trades that wouldn’t be taken by the system after analyzing all the data in the evening. Those trades (entered erroneously MOC) would then be “orphaned”, have to be managed manually, probably lead to difficulties in calculating equity by the system, and not be beneficial in an automatic system. Back in August, 2017 I investigated a compromise: let the analysis run in the evening and then take a smaller 1% allocation MOO entry for all trades along with a larger 10% Day Limit order set at the previous Close minus 2%; This would avoid the problem of orphaned trades and appears to hold promise, but would need to be investigated in more depth. Here’s a link to that post in the OV forum with much more detail on the subject; https://www.omnitrader.com/currentclients/omnivestforum/thread-view.asp?threadid=7845 Further down in that thread, I analyzed using no MOO orders, but just 2 Limit orders; Fewer trades and better results, but like I said - this would need to be tested some more, and the implementation might be difficult. I’m also not trying to detract from what you’ve documented so well, Steve, but can I suggest some further investigation into your results to see if we can extract more information? Your Market Orders info shows a MDD of 75.4% and 4,289 trades. Your MOO Orders info shows a MDD of 88.3% and 12,622 trades (considerably more trades). Maybe there’s something going on that's not obvious from the charts…??? Anyway, I hope this discussion will yield some more information as people (hopefully) respond with what they’ve experienced trading with different order types. Mark [Edited by mholstius on 11/27/2018 11:22 AM] Attached file : MOC v MOO results.png (76KB - 1516 downloads) | ||
| mholstius Veteran Posts: 175 Joined: 1/13/2017 User Profile |
I forgot to mention that the exit results might be improved considerably if the system had a profit target that was filled as a limit order, rather than just signaling an exit to be filled MOO the following day. Mark | ||
| sgbyrne Member Posts: 10 Joined: 1/2/2004 Location: Humble, Texas User Profile |
Your observation on the number of trades for Market and MOO is very interesting. I will run the simulations again to see whether I have slipped up on the settings. Steve | ||
| Ken Wilsdon Member Posts: 6 Joined: 4/12/2006 Location: Calgary, Alberta, Canada User Profile |
First, thank you for quoting me in your first post above. It is nice to know that I have said something that is memorable to others. I am more convinced than ever that what I said then is still true today. I agree with you that MOC orders are better than MOO. Professional money (ETFs and some mutual funds that need to adjust their portfolios daily make up the bulk of this - I have heard figures over 40% of the total market value of the stock market is made up of these pros) normally trade in the last hour or so of trading. The closing is often volatile because lots of money is changing hands to determine fair market value. At the open, to the contrary, market makers are trying to balance the overnight buy and sell orders, and secure a small profit for themselves by scalping. Professionals do not tend to enter at the open, but amateurs tend to make decisions overnight to buy or sell, and those orders are on the books. As Mark has put it so well in the Improving RTM Results With Limit Entries thread, most people who use MOO are basically saying "I trust you’ll give me whatever price you think is appropriate". Let's see if I can move the discussion along with a suggestion. There is a way to get the MOC price at the next open, at least some of the time. That is through limit orders. Place a limit order after the close at the closing price of the day. That way, you have no problem with running the ToDo list before the close; you have all night. And you have a fixed price to buy at, or perhaps even a better price if the market gives you a firesale price. Mark, I noted with interest your table from the Improving RTM Results With Limit Entries thread in Post #36860. While in that post you were interpreting the values of the 1% and 2% below previous close, it also provides indirect statistical data showing the value of using a limit order instead of a MOO order. I have highlighted in blue the section with 0% below prior close. As you mention there, this includes IB commissions.  For FTM, there were 4091 trades that opened less than 0% below the previous close vs 9364 using MOO. That is about 43.7% below, and 56.3% at or above the previous close. From that stat alone, it should be obvious that you will get more bad fills than good ones in your favor using MOO. No wonder so many people lose money in the markets. Let's buy high and sell low. The ARM4 is about 45.6% below and 54.4% above, respectively. PPT and HR are higher. Note that if the price at the open is not below the previous close, but then goes below the previous close, the PPT and HR is lower, but still above the MOO values. Overall, it seems worthwhile to ONLY buy if the price is below the prior close. This makes sense, as if the open is higher than the previous close, then it must go through the previous close (which is breaking a support level, indicating weakness), and then must travel above the previous close to make a profit (this time breaking a resistance level, indicating strength). Having day traded a bit, I know the previous close is a support/resistance level. Doing this round trip down and up uses up some of the "potential energy" for future profits. But sometimes the market makers want to "jump" a support/resistance level so it will trigger the resting stops and clear the books, and they do not have to deal with those levels during the market hours. That's one reason why the PPT and HR are so good when the market gaps down the 1-3% at the open and reverses. While I have dabbled in day trading, I prefer swing trading, and like that over RTM methods. If I were to use RTM a lot, I think Mark's method in ATM M&M and FTM makes a lot of sense, as you are using the best RTMs for the market state where they work the best, and ignoring the rest. I personally feel many RTM systems are like catching a falling knife. They are going down when they buy, rather than even one day of going up to confirm the direction, like you can do in swing trading. There is no doubt you can make money with RTM methods, but they must be really accurate and don't allow much leeway for error. Mark, what would be interesting to see would be the first chart in Post #45704, but using below prior close day after buy signal instead of MOC day of signal for entries. I would expect an increase in HR and PPT, and would be interested in the last columns to see if there is an improvement there. If there is, then perhaps a suggestion would be to replace a straight MOO entry with a limit at the previous close. Then, if you bought additional shares at 1% or 2% below the previous close, it should improve the entire system significantly. Many of these major drops, especially at the open, turn out to be reversal candles, or if not at the open, an exhaustion signal on the downside, indicating the bears are finished and the bulls will take over shortly. Ken Wilsdon Attached file : BelowPreviousClose.png (69KB - 307 downloads) | ||
| sgbyrne Member Posts: 10 Joined: 1/2/2004 Location: Humble, Texas User Profile |
My previous post showed stats for Custom Universal and % of Equity that's why the number of trades were different. Just as a matter of interest I Have performed a couple of simulations with the same strategy but with a different stop. I ran both Market and MOO simulations, see below. I think I may have a liquidity problem!!! Steve [Edited by sgbyrne on 11/27/2018 11:59 PM] Attached file : ATRKingMarket27Nov18.jpg (195KB - 1453 downloads) Attached file : ATRKingMOO27Nov18.jpg (228KB - 1393 downloads) | ||
| sgbyrne Member Posts: 10 Joined: 1/2/2004 Location: Humble, Texas User Profile |
Many thanks for the comprehensive response. I was trying to find some member feedback om 'Market' versus 'MOO', having experienced some issues with strategies I had uploaded to Omnivest. When I read your previous comments it became clear as to exactly what the problem was. I have come to the conclusion that I need to trade up-trending stocks in Omnivest to avoid the big drawdowns. I find that a good list can be updated weekly in a historical list. I have tried dynamic lists but the results aren't as good. I obtain my lists from searches in VectorVest. Having said all that, Mark's and your limit order option is definitely worth further investigation, but as far as I am aware, Omnivest only allows next day 'Market' or 'MOO'. Perhaps Nirvana (Barry) could advise on the limit order possibility. Steve | ||
| mholstius Veteran Posts: 175 Joined: 1/13/2017 User Profile |
Obviously, there’s something going on other than simply MOC vs MOO - and I certainly hope it’s legitimate. If it is, you’ve found the holy grail. If you haven’t already, I’d recommend that you export the trades for both runs to see what’s happening. From what you’ve posted, it looks like you may be taking 1 trade at a time at 100% equity and allowing your ATRKing strategy to do its magic. Is it entering and exiting on the same day? It’s showing 5,730 trades and 80%HR for your Custom Universal method with a 27% MDD. I estimate there should be about 6,300 trading days from 11/19/93 to 11/27/18, so that may be 1 trade / day? What does it show for the % Of Equity method? The % Of Equity is an incredibly straight line with only an 11% MDD. That could be excellent (beyond excellent?) if you’ve actually found a way to enter at the open and then get out the same day with a profit… or it could be a bug. Hope you’re right, but…??? Mark P.S. Write to us after you move to the South Pacific island you buy... [Edited by mholstius on 11/28/2018 7:31 AM] | ||
| mholstius Veteran Posts: 175 Joined: 1/13/2017 User Profile |
I’d thought of doing the limit entry test using the 48,000 trades too. It may give us some better statistics. My wife & I are about to become grandparents again this week. Between that, the holidays, and a couple trips, I don’t expect to have much free time until the New Year. Sigh. Do you have any good short-medium term non-RTM strategies you’d recommend? Mark | ||
| FG Member Posts: 39 Joined: 1/8/2004 Location: Atlanta, Ga User Profile |
Trade"? Thanks, Fred | ||
| mholstius Veteran Posts: 175 Joined: 1/13/2017 User Profile |
Maybe someone else knows the answer...??? Mark | ||
| Hafnium Regular Posts: 74 Joined: 1/1/2004 Location: Fairfax, VA User Profile |
You stated the following on MOC orders "Given all that, there are some serious practical problems with MOC orders in an autonomous system" I have found that autonomous MOC orders actually work fairly well (11 months of live trading with OT2018/ATM/Macro Micro Steps to implement -Modify the tradeplans in use to use MOC orders. -Set the Autotrader to use forming signals only. -With GXtrader/IB - MOC orders must be submitted before 3:45 EST. Since my normal OT run takes 22 mins (250 Liquid stocks) - I set my AutoTrade Run Time to start at 3:20. The orders are generally submitted at 3:42 , just prior to the 3:45 deadline. -Every now and then, the final market closing price will move enough to cancel a trade entry signal (as the entries are submitted before the actual close-this happens roughly 1 or 2 times a month). This is not a problem logistically. Even though the strategy entry signal disappears from the vote line the actual trade remains active - the strategy trade plan manages the trade to its full completion. These trades have the same overall characteristics as the normal trades (HR, PPT, etc...) LD | ||
| SteveL Veteran Posts: 262 Joined: 8/19/2005 Location: Boulder, CO User Profile |
I concur with LD. MOC trades work fine as he has described. Personally, I've changed from using MOC to MKT orders, otherwise just as he described. With MKT orders, if perchance the analysis takes a little longer than usual, the order (entry or exit) will happen at open the next day. The analysis is based on data that is 30-40 minutes old and some trade signals may disappear, but the actual trades are comparable in performance to the PortSim trades. | ||
| mholstius Veteran Posts: 175 Joined: 1/13/2017 User Profile |
The MKT order sounds like a good idea, allowing it to finish the run after the IB limit and prior the close with the AM fill backup. Mark | ||
| sgbyrne Member Posts: 10 Joined: 1/2/2004 Location: Humble, Texas User Profile |
I was skeptical too! I have run dozens of simulations and matched sim trades with the charts. I ran a single symbol sim AMGN from 2010 to 2018 and a Nas100 sim with the same timescale. The spreadsheets and graphics are below. I intend to install Omnitrader 2019 tomorrow and rerun the simulations. I'll wait a while before I rename my strategy The Holy Grail. Steve [Edited by sgbyrne on 11/30/2018 1:55 PM] Attached file : Market2010to2018Nas100CustomUniversal.jpg (162KB - 1128 downloads) Attached file : AMGN.jpg (200KB - 1101 downloads) Attached file : Market2010to2018AMGNSymRun.jpg (163KB - 1097 downloads) Attached file : Market2010to2018AmgnSymRun.xls (102KB - 256 downloads) Attached file : Market2010to2018Nas100CustomUniversal.xls (238KB - 257 downloads) | ||
| mholstius Veteran Posts: 175 Joined: 1/13/2017 User Profile |
Those are some truly amazing results! And without using margin, too… You’ll probably want to make sure you can get a good satellite connection on the island you buy. ;-) On a more practical note, I’d be happy to plug your trades into my extensive analysis spreadsheet if you’d like (no need for any details about your system, just the trades). I suspect the stats would be amazing. Just send me a PM if you want to. (I’ll be gone starting next week). Congratulations, Mark | ||
| mholstius Veteran Posts: 175 Joined: 1/13/2017 User Profile |
Just saw that you'd attached the excel files for the trades. I'll try to get the analysis done tomorrow. Mark | ||
| sgbyrne Member Posts: 10 Joined: 1/2/2004 Location: Humble, Texas User Profile |
Just noticed a minor error in the 'Days in Trade Col' in Market2010to2018AmgnSymRun.xls, I corrected it and replaced. Steve | ||
| gary Member Posts: 16 Joined: 12/13/2003 Location: Houston, TX User Profile |
have you executed this method on the right edge? If so, please show results of that activity. If not, please do it for a few weeks and show those results. If it matches portsim results, please put it in elite trader. | ||
| sgbyrne Member Posts: 10 Joined: 1/2/2004 Location: Humble, Texas User Profile |
Thanks for your interest. I am working on the analysis of 'Market' and 'MOO' trades including the right edge. The results I have shown are for 'Market' trades. 'MOO' trades are totally different with different results. Omnivest has Elite Traders but sell trades are the following day after the exit signal so the results are a lottery unless you have a list which has a high probability of trending up. I am working on a strategy that only selects robust trades that are trending up. I will let you know if I have any success. Steve | ||
| Ed Downs Member Posts: 11 Joined: 2/12/2013 User Profile |
Steve, adding your ATM method to Elite would be awesome. We are able to run ATM methods on machines outside OmniVest, and orders can pass through to the broker during the session. This is how we ran the "Trading Room" several years ago, and allowed users to follow the traders in the room in Real Time. I will have to check to see what the hurdles are to reviving it, but I agree with Mark - if these are actual results, or even close to actual - we'll figure it out! | ||
| sgbyrne Member Posts: 10 Joined: 1/2/2004 Location: Humble, Texas User Profile |
That would be fantastic if we could load up to a separate server and trade in real time or close to it. There would need to be a strategy run immediately after a trade closed out in order to select the next best candidate in ATM. I will be testing my ATM method to see how it performs at the RE over the next few weeks. Steve | ||
| Jim Dean Sage   Posts: 3022 Joined: 9/21/2006 Location: L'ville, GA User Profile |
| ||
| Ed Downs Member Posts: 11 Joined: 2/12/2013 User Profile |
|
|
|
|
|
Administrator
Forum Moderator |
Registered User
Unregistered User |
 E-Mail this thread to a friend |
Toggle e-mail notification |
