Dustin’s Journal

[Anonymous]

Hey Everyone, I have been reading through some of your journals and have really seen the benefit in tracking thought process, challenges, ideas and overall progress… and if nothing else just to blow off some steam in tough markets So I thought I would kick off with mine today on the back of a horrific day yesterday.

Just as a general update, I have made it though all of the material and feel pretty comfortable with the process and coding. I started off focusing on NASDAQ and S&P500 relative momentum strategies. By the time I finished these the market was choppy and near the index filter so I never did implement these, however, I am now ready to turn them on once the market turns more positive (whether that is in 2019 or later, who knows).

For the last 2 months I have focused on mean reversion systems. I developed 5 systems that I am currently trading. This sounds like a lot but in reality it is more like 3 systems as there are 2 systems that are quite similar and correlated and another 2 systems that are similar and correlated – often positions overlap. I guess I like the diversification of running more systems and I was struggling to decide between them. Four of the systems are MOC and one is exactly the same as a MOC system but with multi-day hold.

I have been slowly scaling up/introducing systems over the past 6-7 weeks. Its been a fun process – there have been some ups and downs but overall the systems have been performing as they should. One of the five systems has really been performing poorly but not outside of any normal drawdown so I won’t think about changing anything there before 6 months of trading or a drawdown far outside of historical norms. Prior to yesterday I was up ~1.5% on the account from all of the systems – not for the sake of evaluation but just nice to see it working.

Probably like many others, my systems were blown out of the water yesterday. The account took a 4.2% drawdown in one day yesterday… OUCH!! Obviously not fun but nothing outside of my backtests or normal one-off bad days. So back to placing the trades tomorrow since the market is closed today.

In terms of system development, I have struggled to find any trend following strategies that I feel as comfortable with as the relative momentum strategies. I am playing with the idea of combining relative momentum and trend following in the sense of trend following type rules and a subset of a broader universe based on relative momentum rules. It will take some time as the only way seems to be to use StaticVarGenerateRanks which complicates the coding. Something to pick away at though and I am happy to share as I progress it.

[JulianCohen]

One issue with developing a trend following system after an MOC system is that the results can look weak in comparison. The CAGR is only around 20% and the MDD is often 25+%, even into mid 30’s depending upon how far you look back. I have a WTT system that seems to go for a couple of years with very poor results and then three or four with great results. I trade it to give me some variance in my equity curve. It is only about 10% of my portfolio but on the good years it will be great and on the bad years it won’t matter too much.

[Anonymous]

Thanks Julian. Ya, the MOC systems definitely seem to provide more upside potential. The way I have thought about it is that I’m not sure a) if anybody is actually achieving the +30% CAR results that MOC systems appear to provide, and b) when I think about +25 years, I think the relative momentum systems are much more likely to maintain their performance edge. I would love to hear if you and others have been able to achieve MOC returns in line with the super promising backtests of +30%.

It also makes me a little bit nervous that the super performance of MOC systems usually requires significant leverage. The majority of MOC ideas that I have tested seem show performance deterioration for the last few years – who knows if that is a structural change (i.e. machine learning and data mining slowly finding and killing the systems) or if it is cyclical (i.e. low volatility in recent years, which may change in the future). Given the leverage requirement and the performance deterioration, I am not as confident that those high returns will be achieved going forward… time will tell.

From my perspective, realistically if I were to compound my account at 15-20% for a very long time, I would be ecstatic with those results so I will still look to trade the relative momentum systems as a meaningful portion of my capital. I suppose if the MOC systems start crushing it then I may revisit the capital allocation!

[ScottMcNab]

The MRV systems with hold times of 3-5 days seem viable alternatives even without any leverage Dustin…gets you into that 15-20% range and often the returns have low correlation to the rotational momentum systems. I used to be 4x MOC but use 2x MRV now along with rotational momentum systems….whatever lets you sleep easy…go with your gut ….”(insert favourite cliche here”)

[GlenPeake]

Well Done Dustin on creating the 5 systems!!! Nice stuff!!!

If it’s one thing the Mentor Course does is open you eyes to the different possibilities, building those robust trading systems and trading the way professionals trade.

In regards to the MOC systems and the 30-40%+ returns, as the CBT code has only been available/live the last few months, I guess will have to wait and see what the market does. However as you say…. compounding 15-20% year on year would sound pretty sweet to most I think.

[Anonymous]

Thanks guys for the thoughts and comments. Scott I will definitely spend some more time on multi-day MR systems going forward. I think its more likely that those systems hold up longer term as well given that some traders will not be as willing to hold overnight and multi-day. At 2x leverage, at least it feels like it would be less likely to completely derail for whatever reason.

On another note, what a wild ride the last 2 trading sessions has been! At one point today I was starting to get that uncomfortable feeling in the pit of my stomach (yes – I have been looking at the market too much in these early days with my systems!). Then to see a complete reversal and make back effectively all of the losses from Tuesday. Certainly a a good market to test my resolve on the systems and illustrate how it is best just to trust the system and place the trades. Had I been trading discretionary I certainly wouldn’t have been loading up my account like my systems did today! Anyway, very interesting trading days!

[Anonymous]

Thanks Glen – ya, after reading many of your journals I feel really spoiled to just have joined recently and be able to benefit from all the work and resources that all of you have contributed. So a big thanks from that perspective.

[LEONARDZIR]

Justin,
Congratulations on your progress.
I wouldn’t be sweating a 4 % drawdown if I were you. It is guaranteed that your worst drawdown is ahead of you and will have you questioning whether your system is broken. Until October I had been cruising along with a good profit in my US mrv system and rapidly went down 22% in October before the system went flat. My aggressive Nasdaq system had a 36% profit for the year and by the end of October was up 6% for the year. Similar issues with my rotational systems. However all the drawdowns which were considerable were consistent with backtest results. My view is that success gets down to psychology and how you view the meaning of money in your life. I think developing the systems is the easy part. Good luck on your journey!

[Anonymous]

Thanks for your thoughts Len. I should clarify that the amount of money I am trading in the systems now is not what would make me feel uncomfortable. Its more the time and effort that has gone into it all that makes me really want it all to work. I think seeing the sharp moves down just made me think… maybe I over-optimized the system and the system will flop in real time. Reflecting on the past few weeks, I think I just need to trade through a few more of these wild markets in order to see how the systems respond and go through drawdowns. To your second point, ya, I have been reading about meditation and the benefits for traders. Probably a good habit in any case and if it can help put money and trading into perspective then its a good secondary benefit.

[Anonymous]

Hey guys, I was thinking about backtest metrics to review and optimize for and wondering if there isn’t a metric that helps one understand the consistency of the system. More specifically, I often am looking at CAR/MDD for optimizing but then when you look at the individual backtests, a substantial portion of the returns are generated in a small number of periods. I would prefer a system that shows more consistency and is less ‘lumpy’. The metrics that seem to help understand this are Standard Error, Risk-Reward Ratio and the K-Ratio. I question if optimizing for one or any of these ratios with secondary objectives of CAR/MDD and CAR wouldn’t be a better approach in terms of having systems that are likely to perform in the future as they have in the past. Would love to hear any thoughts.

[Anonymous]

And another question that I would be happy to receive thoughts on is how much weight you give to pre-2004 returns when evaluating systems, specifically mean reversion systems. I was listening to the Cesar Alverez interview on BST and it does seem logical that the market has changed so much in terms of quant involvement and market structure that anything before the mid-2000s seems somewhat irrelevant for high-turnover systems like MOC and short term mean reversion.

I am specifically thinking about it because I have been retesting one of the systems that I am trading and, rather than optimizing for a historical period, I optimized for the last 3 years. To test the system robustness, I then extended the OOS period backward. Performance is clearly much better in recent periods (even much better going back to 2010), but as you would expect bigger drawdowns are introduced. However, the big drawdowns are really in 2001-2004 and for no apparent reason (i.e. no market crash etc) and the system does just find during 2008, the 2011 volatility and the 2016 selloff. I am really tempted to trade the more recent parameters given that the system weakness was so long ago.

I would add that I kept the Oct/Nov data as OOS data for testing post-optimization and the drawdown for the updated parameters was 18% vs 14% for the old parameters. So perhaps there is more risk with the new parameters but the returns in recent years seem to more than offset the added risk.

The returns and drawdown tables are attached. The one with large drawdowns in 2004 is the one with parameters updated form the most recent 3 years.

[SaidBitar]

Regarding short term trading while backtesting i do not go so much back most of the time i back test from 1/1/2007 the duration is long enough for short term to have enough number of trades and it covers bear market and bull market
i think in the early 2000s not sure exactly when in US they used to use fractions and not decimals so the back test result will be somehow not accurate.
Regarding IS and OOS honestly i am not big fan of OOS and when i do optimization i tend to take not the best value but the middle value in one area where all the results before and after are relatively the same and the best test and OOS is real trading.

[RobGiles]

You might find this an interesting read Dustin

[img ]https://enjoytherideworld.odoo.com/slides/slide/the-etr-comfort-ratio-a-better-way-to-measure-return-to-risk-21[/img]

[Anonymous]

Thanks Rob – that is going on my list of ideas to try to code up!

[Anonymous]

December was the first month where I had all of my systems live in the market – perhaps poor timing

Dec 2018

MOC/MR Portfolio (US): -8.46%
Nasdaq Rel Mom: 0% (cash)
S&P500 Rel Mom: 0% (cash)

Wishing a great 2019 to everyone!

[Author]

Posts