Quantopian's community platform is shutting down. Please read this post for more information and download your code.
Back to Community
Meb Faber's Ivy Portfolio

Meb Faber's book Ivy Portfolio describes a very stable strategy befitting of an infinite investment horizon institution.
The idea is that you have 5 asset classes
Domestic stocks (SPY)
Foreign stocks (EFA)
Commodities (GSG)
Real estate (VNQ)
10 year treasuries (TLT)

Each take up 20% of the portfolio at any time. In order to keep a place in your portfolio, the asset class's price must be above its 200 day moving average. Here is the results of the backtest when you allocate assets under their 200 day average into cash.

Ie if SPY is the only asset class above 200 day mavg, the portfolio would be 20% SPY, 80% cash.

4 responses

Here is the backtest when nocash is set to true. This means that if SPY is the only asset class over its mavg, you are 100% SPY.

The first backtest has much lower volatility. Why don't we see what happens when you use nocash=False and increase leverage to 2.

It works better! I wonder what the mathematical reasons for this is.

Some Q members have said that IAU (gold) may work better than TLT because it has negative correlation with the stock market. Here is the code with IAU instead of TLT, max leverage of 1, and nocash=True

Keep in mind that IAU doesn't exist until later on in the backtest, so these results may be skewed.

Let's datamine! If IAU is better than TLT, lets see what happens when you increase leverage and set nocash=False

Opps :( This leads to half a decade of sadness.