Lets start with the oldest store of value there is... gold!
Prefer cashflow generating assets? How about some long-term US treasuries!
Like to hold a more diverse basket of assets both long and short, and use trend to determine positions? How about the Mt Lucas index net of 85bp to represent $KMLM?!
Another trend based index from AQR to get a bit further back in time.
Too boring and you need more risk? Okay I doubt anyone really wants this portfolio but how about adding 2/3 stack of US small-value!?
75% drawdown, woof!
None of these backtests are intended to represent actual returns you can expect moving forward, and in some cases not even what you could've got in the past (no trading costs in KMLM index, no costs at all in AQR), but they do show the potential relative utility of UPRO (cont)
In fact, in all cases it outperformed a traditional margin loan with a borrowing rate 2% above the 3-month t-bill, not bad at all, and you can even implement this in a retirement account! (cont)
Does rebalance timing matter? Yes, but that is true of any portfolio, levered or not. Not shown, but rebalance bands may be most appropriate.
Last but not least, here is monthly rebalance between 33.33% UPRO, 66.67% 3-month t-bill, and -100% S&P500 to show the rolling drag.
Someone recently said that manage-futures/trend products without equities (KMLM for example) would not protect as well in an equity drawdown (no equity shorts) so I decided to setup a little backtest to demonstrate that this at-least historically has not been the case...
I started with a simple dataset from AQR that uses 12-month lookback, rebalanced monthly, and has separate returns for long/short trend on equities, commodities, bonds, and fx.
I then used risk-parity weighting and made a monthly rebalanced portfolio of all 4 asset-classes, as well as one for the 3 without equities.
@jagmavi@rcmAlts@BillBrewsterTBB After deep diving with every moment of free time... I came to the conclusion that I am most interested (and convicted) in a pure trend following approach, and don't want to dilute it with carry, value, or other global-macro type strategies. This is because I want to... (1/x)
@jagmavi@rcmAlts@BillBrewsterTBB ... maintain a relatively small allocation (12.5% today, growing to 20-25% long-term) to damp poor equity periods, and don't care about absolute Sharpe of the CTA portion, so I want convexity. For this same reason I also want as volatile of products as possible, fee-adjusted (2/x
@jagmavi@rcmAlts@BillBrewsterTBB ... I do worry about single-manager risk though so I wanted 3 products. I chose $EQCHX because I like Jerry's pure process, no volatility targeting (let winners run), and no cute additions. I chose $KMLM because it is lower cost to reduce overall weight, but still has a (3/x)