Tim’s Journal

[Nick Radge]

Matt,
You need to look beyond the MR systems helping ‘every day’. Don’t micro manage or micro assess. Step back, look at the results during sustained down periods when the trend systems are fully off, i.e. 2008.

[TimothyStrickland]

Nick remembers me sending him emails about my (broken) MRV system last year. I thought it was broken, we went back and forth several times and couldn’t find anything wrong with it so I just stuck with it.

I’m doing good lately but tomorrow someone else will.

[TrentRothall]

Great stuff Tim

[Anonymous]

Tim,

It sounds like your system is similar to mine in some ways, but different in others. Sounds like you are still measuring recent history via the ATR, then issuing a limit order for a buy stretch down based on that ATR (I only use a 1x ATR I don’t use a 2 or 3 or 4 times ATR like some other people do, simply because I am trying to aim for lots of signals and lots of fills). Perhaps you get a fill within that day, but then on the sell it sounds like you are not doing the Market On Close order within that same day of filling (like I am). You are instead waiting for the close, doing your assessment/exploration after hours of the price movement within that days’ session then issuing a Market order to sell on the next open should any limits/triggers/indicators you are running have been breached on any positions you are holding within that day (whether they are symbols held from previous days trading, or held from getting a fill on your limit order within that same day). So I am guessing you would be running with (at most) 2x leverage and holding always for a minimum of overnight? It must stink having to face the overnight opening gap. It is clear that gapping is becoming larger, which stinks.

I don’t bother with an index filter on my system. I trade it every day. Every time I have tried to add a filter the results get worse across the board. That said, the primary reason of me running this day trading MOC style system is to try and grab those intraday swings down (and then hopefully high again) so I do need to accept volatility. However the issue I am having psychologically at the moment is that if this system isn’t going to perform much in low volatility, boring, upward trend markets when the trend/Mom systems are firing along, and if it is going to falter somewhat in these heavily volatile times, then this means it is likely going to only be firing when things are rolling along at a “Not too little but not too much” level of volatility. What is not too little and what is not too much? Who knows, but it is my personal opinion that the speed and depth at which the market whips around is going to be here for some time to come, so how I do I find the right level of settings/comfort to trade it right in the butter zone, to stay out when it is killer levels of volatility, stay in during medium levels and don’t get too excited or expect too much when vol is low?

I know none of us have all the answers, just thinking out loud. Maybe I’ve already answered it myself and just need to add a simple market volatility filter and have it drive position sizing or margin such that it allocates most when market is in the volatility butter zone, then scale it back somewhat with either low or extremely high volatility. The problem I have with this is that adding measures like this make me worry I’m adding too many dimensions to the recipe and potentially causing a curve fit.

[ScottMcNab]

I found that combining a MRV (ave 3-4 day hold) with a MOC (even if exactly the same system except for the exits) often produced a smoother equity curve..psychologically easier to trade too..
having mrv and moc both at 2x also reduced my stress levels back when I was running them together…

[ScottMcNab]

Just did some quick tests Tim…can run them both with no leverage at all and get > 10+% CAGR you are aiming for
since 2010, the MRV I use on RUI at 10×10 did 20/-11;
same system dropped straight into MOC over that time did 13/-4
combined they hit the targered CAGR you’re after with maxDD in single figures (and a smoother equity curve that either system alone)
Does a 401k have limitations regarding trading frequency?

[TimothyStrickland]

For the most part Matt yes. My system exits the trade on the following day, which is probably why my gains have been decent lately. Much of the gains in the previous months have been during the overnight session.

Matt, are you running multiple systems? In my opinion (and it is just that I am no expert) running multiple systems would probably be the better option. Maybe create 3 MRV systems, one to handle low volatility and then have an index filter on it for large volatility times. Keep yours like it is and maybe make some small adjustments and then have one similar to mine that really produces during times of high volatility. Just some thoughts, I am sure others can comment on that. I am sort of doing that my momentum system and MRV, my MRV performs well during high volatility while my momentum system does better in sort of a mid-range performance. I don’t have a system yet for low volatility times.

I think Nick says the closest thing to a holy grail is having multiple systems. At some point, I want to be running 5+ systems once my capital allows.

[TimothyStrickland]

Thanks Scott. I was going to make a MOC but just haven’t gotten around to it. My wife and I are in the middle of changing careers. Thanks to Nick and Craig, once I started coding in Amibroker several years ago, I fell in love with it and now I am getting a degree in Computer Science and transitioning into a Software Engineering position at Boeing (DevOps), I LOVE to write code lol. This course literally changed my life, so much so that my wife just quit her job and is going to school full time to be a software engineer as well. Crazy

My 401k limits trading frequency to every 15 days or I incur a penalty. Additionally, it limits what you can invest in, so I cannot run a momentum system in it and I get calls all the time from the person who runs it asking me why I want to move the money, that it is better to stay in yada yada, typical ignorant sales person talk that doesn’t understand how the market works. I just respond with if you can make me 30% returns this year I’ll stay in, they usually don’t respond.

The 401k does have something called the S&P 500 index fund…but it doesn’t match the S&P exactly and I have backtested just being long the index with and without an index filter in many ways and found that the difference is not worth the effort to move the money every month. The reason why the index filter is so effective is for aggressive systems that are capitalizing on a few stocks with stronger moves, you don’t want to then hold those stocks as they drive down the market. When I am long the entire index I lose that large upward momentum and with that the CAGR, so running an index filter doesn’t really help.

I really like the idea of the momentum systems, I don’t have to watch it every day and just put my trades in for the month and walk away. My emotions usually get out of wack watching my MRV, I have gotten used to it now but still… My current momentum system has a CAGR of 32% which is very nice but it is only long a few stocks which make it susceptible to a black swan event, which I can’t afford with My wife and I’s life savings. What I’d like is something in between which Nick gave me some ideas and maybe the MOC MRV setup Scott mentioned. Something that captures better gains than the S&P but is not so exposed as my current Nasdaq momentum. I am testing some ideas now.

[ScottMcNab]

Just making stuff up now on the fly but….based on restrictions and preference for momo…what about option of a system that keeps large portion of 401k in cash and rest in momo…maybe a SPX system that has 20 positions at 3%…it may hit the 10% cagr with an acceptable (<20%) maxDD ?

Dont know how to test it but another possibility would be 20 positions at 4.5% with other 10% in otm puts for some insurance… that is a backtest I’d like to see…something simple where just bought put on 1 Jan and 1 July each year for example…would need someone with a passion for coding to explore

[TimothyStrickland]

That is an interesting idea. Does Amibroker allow the code to do 50% cash and 50% vested, I can try it. I think last time I did that the code doesn’t allow it, it assumes all cash is vested. Your right, I think I would have to tinker around with the code quite a bit.

[TimothyStrickland]

This is why I mentioned my gains would likely be short-lived. The market will open in at least one limit down, probably 2 wiping out most of the gains I made on Friday, I may actually have taken a loss.

I ran my MRV explore this morning and got nothing, I checked to make sure everything was good and apparently all stocks that could have been hit are below their 200 day MA so my system is not taking any more trades, and signals to get out of all trades it is currently in.

[TimothyStrickland]

I have spent more time reading and studying trading lately after this unwinding. I thought I knew some things and found out I did not. Failure is the best teacher. It is easy to get complacent in a bull market and all your accounts are at an all-time high, however, when your trading accounts need you the most is in an uncertain volatile market with lots of confusion.

Good article if anyone is interested.

CORE RISK MANAGEMENT: Using Key Price Levels & MAs to Avoid Major Downtrends

[GlenPeake]

Thanks Tim…. good article, some sound advice in there.

[TimothyStrickland]

This post is mostly for Nick or anyone else who wants to critique my ideas. During this downtime, I have had some time to think about a lot of things. As I said earlier I am working on some ideas on where to put my money (not in the 401k) here are my thoughts:

I am thinking about allocating 50% of after-tax capital to my NDX momentum system, the aggressive one that gets 32% CAGR on the backtests.

The other 50% was going to be run in the futures market which would long the S&P 500 (ES) there. and step out on break below the 200-day MA or 10 month MA that I have tested. There are a few reasons I chose futures. The futures are taxed at a lower rate than stocks, which would improve the performance of the system tremendously, over the long term it would be a full 2%. This is meant to be a somewhat conservative system that matches the index with the tax advantages and prevents large drawdowns by going to cash in bear markets.

My 401k would be my conservative fund, I get a 100% match from my company up to 8% of my income pre-tax so it is about 16% of our income going to it every year, it would stay in something like a govt bond fund which is extremely conservative and chugs along at about 2.5%

I then plan on borrowing from my 401k like I do every 5 years and seed my MRV with that money for my short term system.

I tried backtesting other methods a little but the NDX momo meets my Drawdown and CAGR requirements if the other chunk is in some conservative area. I’ll have to do the exact math to see what the actual max dd I would experience and may have to put more in the 401k if I feel the risk is still too high.

My question is, anyone tried doing anything like that in the futures market? I would need to calculate the risk etc like everything else but it seems like a good idea. Any other comments?

[TimothyStrickland]

MR HFT: -6.5 % not a lot of trades this month.

Momentum Systems:

NDX Aggressive: +3.6% in cash now.
ROR: +10.3% in cash now.

Moving Average Systems:

SPX: +0.7%

The momentum and MA systems had gains in march because of the first-day rally.

[Author]

Posts