Looking for Portfolio Feedback and Modeling Advice from Experienced Return Stacked Fund Users
Hello everyone! I wanted to try an experiment today and see if I could emulate one of my favorite global portfolios leveraged up to 2x using the suite of return-stacked products currently available.
For reference, the target allocations I’m shooting for (by notional weight) are:
20% US Equities
30% International Equities
20% Bonds (avg dated maturity 5-8 yrs)
10% Gold
10% Futures Trend
10% Commodity Carry
These weights were born out of Robert Carver’s book “Smart Portfolios” by taking the maximum geometric mean and maximum Sharpe portfolios and creating a compromise portfolio that accepts the potential embarrassment of using the maximum recommended allocation of ~30% to ‘alternatives’ such as CTA, MF, carry etc.
Using return stacked funds, I came to the following allocations to get as close as possible to the stated portfolio after accepting that the international exposure would likely be reduced and the bond arm would likely be enhanced. At 2x leverage I think this is a fine compromise given the lower volatility of bonds in general compared to equities. Here is the portfolio I came up with:
25% RSSB (60% US Stocks, 40% International Stocks, 100% U.S. Treasuries)
20% RSIT (100% Developed International (ex-U.S.), 100% Futures Trend)
15% RSST (100% U.S. Large Cap, 100% Futures Trend)
10% RSBT (100% U.S. Treasuries, 100% Futures Trend)
10% RSBY (100% U.S. Treasuries, 100% Futures Carry)
10% RSSY (100% U.S. Large Cap, 100% Futures Carry)
10% GOLY (100% U.S. Corporate Bonds, 100% Gold)
Combining these together gives the following notional weights for the entire portfolio
40% US Equities
30% Developed International (ex-US)
55% Bonds (45% Treasuries, 10% Corporate Investment Grade)
45% Futures Trend (blend of pure trend and CTA, assuming here roughly 30% trend and 15% other strategies)
20% Futures Carry (Yield)
10% Gold
So far, I think this is the closest I can get to the target allocations, but I’m sure someone out there with far more knowledge and experience in the return stacked funds space can probably point out where to make a few more incremental gains. The link to the portfolio simulation is down below with some pretty generous assumptions on how I crafted it. If anyone has a better strategy for simulating managed futures, specifically commodity carry, I’m all ears. However, after digging through the forum it seems pretty hard to do in testfolio at the moment without making a lot of assumptions or blending several ETFs together.
U.S. Equities – simulated using VVSIM at 40% notional weight
Developed Intl (ex-U.S.) – simulated using VXUSSIM at 30% notional weight
Managed Futures + Futures Carry – simulated using a split of CTASIM (25%), DBMFSIM (25%), KMLMSIM (15%) <-- Here KMLM got a smaller split as it is pure trend. I made the assumption that CTASIM and DBMFSIM would provide the remaining mix of trend/carry/MF allocation. This is the weakest spot in the simulation in my opinion
U.S. Treasuries – simulated using IEFSIM (45%), chosen due to average weighted maturity typically between 5-7 years
Investment Grade Corporate Bonds – simulated using LQD (10%). Backtest history is capped to 2004 due to CTASIM, so no reason to try and find a corporate bond fund with a longer history.
Gold – simulated using GLDSIM (10%)
Financing cost - simulated using CASHX(-100%) + 50 bp
The estimated financing cost was calculated by subtracting CASHX (-100%) from the entire portfolio and applying an additional 50 bp spread to the financing cost as a conservative estimate of the cost of borrowing for leverage. In addition, I estimated an additional 1% return drag for the expense ratio of the funds themselves. If I were to run this, it would be in a tax-advantaged account so I didn’t feel the need to estimate taxable distributions. I think the backtest looks fine for what it is. I did zero allocation optimization. I just looked at the fund holdings and said I can get close to my targets using the published weights. If anyone has any suggestions on new funds I may not be aware of let me know. I would personally love a VXUS + GLD fund to help up the gold and international allocation and trim some of the overweight futures holdings. In addition, I'm well aware that the backtest overstates the returns because we don't get to simulate back into the 90's and the true CAGR is probably closer to 11-12%. Simulating back to 1990 is another topic I'm going to focus on in the future.