Most of 17 assets in the list are highly and positively correlated.
If you will leave just 2 but negatively correlated assets you will get:
more return,
less draw down,
never been under water.
Correlation does matter.

Yeah, but you've introduced a big wad of future snooping by picking VDC which has done great the last few years. I do not think this is an improvement.

Backtest of an algo which uses these principles. Naturally you can turn up the leverage if you want, and/or add your own forecasts if you have predictors. I might try that next.

Here is another pair without VDC

Sure, you could also pick BRK or IBB or any one of hundreds of stocks that have done well in the last 8 years.

Here is the same thing but backtested by algo on VTI-TLT.including bear and bull market.Not only bull as yours.

Again, the US stock market has done great for the last few years. I feel like I am repeating myself.

As I remember period 2007-2009 was not great for stock market (Bear market).
So my backtest has both legs : bear (2007-2009) and bull 2009-2015).
Here is another backtest : bull (2003-2007) and bear (2007-2009).

Super, thanks.

And bonus: pure bear market 2007-2009 performance

FWIW, your backtests are not true to his system, because you didn't recalculate the diversification multiplier given your new set of assets.

This is yours algo I just changed symbols and if it is not true to his system you can fix it or take it as is.

Indeed, you did. Adding value wherever you go.

Thanks Simon.

Have you ever shared an algo of your own, or do you just spend your time adding cognitive biases to other peoples' and complaining about stability and consistency metrics?

nice!! Thanks Simon (..and Vladimir)

Portfolio Analysis using pyfolio