Len’s Journal

[LEONARDZIR]

Another lesson I learned this week is that you will have a drawdown greater than you see in backtesting. In the last 2 weeks my MOC had a13% drawdown. Since 1995 prior to the past 2 weeks max drawdown was 7.5%. That makes you start to question if your system is broken. Despite a higher Drawdown as I mentioned my MOC is at new equity highs.
Evolving to become a profitable system trader isn’t easy.

[Nick Radge]

Quote:

Since 1995 prior to the past 2 weeks max drawdown was 7.5%. That makes you start to question if your system is broken. Despite a higher Drawdown as I mentioned my MOC is at new equity highs.

Len,
If you have selection bias issues during the last week then that would most likely not show up in the backtest. You should do the manual exercise of selection the worst trades for that period and you should see it.

If you want to send me your code I can show you the drawdown in the testing.

Just on selection bias and testing; I have been doing a lot more work with the Amibroker MCS function and I believe it to provide a much better indicator of outcomes; both on singular days and over extended test periods.

We’ll be rewriting that section of the course to show this functionality better.

Nick

[LEONARDZIR]

Nick,
I will send along the code. Even with having 40 positions there were clearly selection bias issues on the big down days

[LEONARDZIR]

FEB

NDX rotation -0.8%
US MOMO -5.63%
US MOC -0.8%
VTI -3.4%

[LEONARDZIR]

Here is an IB issue i just realized I had.
Last thursday I put in my usual buy limit orders for friday. Two of my positions were filled long thursday night on Island and then reversed. However my API was running and thought I was long both positions and sold them MOC. This left me short both positions over the weekend. There is no compensation because IB sent me a message. Means I have to start checking my API during the day which I don’t really want to do.

don’t want to do.

[LEONARDZIR]

I am reading Daniel Kahneman’s “Thinking, Fast and Slow”. He is a psychologist who won the Nobel prize in economics. He is one of the founders of behavioral economics. One of the big takeaways from his Prospect theory of behavior is how hardwired we all for risk aversion. It probably has some evolutionary benefits. In fact fear of loss is a much more powerful emotion that anticipation of gain.
This fear probably explains why traders fail and dump profitable systems when they go into drawdown.

[Nick Radge]

And the other side of that is that successful traders don’t risk as much as they probably should.

Quote:

No matter what kind of math you use, you wind up measuring volatility with your gut.

—Ed Seykota

[LEONARDZIR]

Nick,
Another takeaway from Kahneman’s book is that in losing situations (negative expectancy)rather than take a sure loss we tend to gamble and are willing to take a bigger loss in hopes that we can be profitable. Whereas when we have the odds in our favor we will tend to take the sure gain rather than gamble for a larger return.
It all comes back to risk aversion.
My takeaway is to plan to increase my investment in my systems when they undergo some drawdown

[LEONARDZIR]

March 2018

USMOMO. -1.24
US MOC. -5.06
NASDAQMOMO. -4.4
VTI. -1.2

[LEONARDZIR]

April,2018

Nasdaq MOMO -2.1%
US MOMO -0.5%
US MOC +0.95%
VTI +0.4%

[LEONARDZIR]

May 2018

US MoMO 0%
Nasdaq Momo 16.9 %
US MOC 0.4%
VTI 2.7%

After running the SB codes found my mr swing on the R1000 had better statistics than my MOC so am dropping MOC and will trade MR. MR has less leverage and less exposure. Both have similar entries.

[Nick Radge]

Nice effort on NASDAQ Momo!

[LEONARDZIR]

Thanks Nick.Wish I had all my money in that system but too volatile.

[LEONARDZIR]

June 2018

US MOMO +1.3%
US MR -0.4%
Nasdaq MOMO -3.0%

Stopped trading US MOC in favor in US MR

[LEONARDZIR]

On one of Nick’s email posts he mentions a number of blogs he reads. One of them was from Charles Bilello of pensionpartners. I went on their site Pensionpartners.com and downloaded 2 white papers on intermarket analysis and asset allocation. Their thesis is that the bond market leads the stock market. They showed that for the prior month if the 30year treasury outperforms the 10 year US treasury you should rotate into bonds. If the 10 year US treasury outperforms the 30 year you should be in stocks. The have several portfolios including all stocks/all bonds, and 3 other allocation strategies based on this rule. Their backtesting shows impressive risk adjusted returns compared to buy and hold.
I am wondering whether the bond signal
would be a good filter for the US rotational strategies.
Also worth pointing out that last month the 30year treasury outperformed the 10 year suggesting a defensive posture.

[Author]

Posts