**My UPRO/BLV Strategy**

A lot of the strategies I write about on Seeking Alpha use leveraged ETFs. One of my favorites is combining a leveraged S&P 500 ETF with a bond fund to achieve net beta of 1, but positive alpha.

For example, in my article A Simple SPY Top-Off Portfolio, I suggested one-third ProShares UltraPro S&P 500 ETF (NYSEARCA:UPRO), two-thirds Vanguard Total Bond Market ETF (NYSEARCA:BND). UPRO is a 3x daily S&P 500 ETF, and BND is Vanguard's total bond market ETF. BND's beta fluctuates, but is generally around 0, and its alpha is typically positive. The portfolio's net beta is 1/3 (3) + 2/3 (0) = 1, and its net alpha is two-thirds BND's alpha (ignoring UPRO's very small negative alpha due to fees).

In general, one can expect the portfolio's positive alpha to translate to better performance than the SPDR S&P 500 Trust ETF (NYSEARCA:SPY), e.g. higher raw and risk-adjusted returns, smaller drawdowns, etc.

Shortly after publishing that article, I decided to implement the strategy for my own retirement portfolio. I started with UPRO/BND, then decided to swap BND for the long-term bond fund, the Vanguard Long-Term Bond ETF (NYSEARCA:BLV).

BLV tends to have negative beta, so I use a 100-day trailing average to estimate its beta and the corresponding UPRO/BLV allocations to achieve net beta of 1. The formula for the UPRO allocation is (1 - beta) / (3 - beta). I do this on the first trading day of each month, and rebalance if the effective beta is outside (0.9, 1.1).

**Performance in 2016**

Rebalancing trades on the first day of the month was required for 4 of the 8 months so far. Figure 1 shows growth of $10k for my UPRO/BLV strategy so far in 2016, alongside SPY and BLV. Performance metrics are given in Table 1.

Table 1. Performance metrics for various funds.

Fund | Growth (%) | MDD | Sharpe |
---|---|---|---|

UPRO/BLV, target beta = 1 | 19.9 | 7.1 | 0.14 |

SPY | 9.7 | 9.2 | 0.07 |

BLV | 15.0 | 2.7 | 0.17 |

Figure 2 shows BLV's 100-day trailing betas in 2016. Its beta ranged from -0.30 to -0.13, and generally became more negative over time.

Considering that BLV's beta fluctuates substantially, I would advocate for the sort of moving-average approach I used here, as opposed to a fixed allocation. However, performance of fixed allocation UPRO/BLV portfolios may also be of interest to readers. Figure 3 shows mean vs. SD of daily gains for various allocations.

BLV's aggregate beta for 2016 was -0.214, which corresponds to a UPRO allocation of 37.8% to achieve net beta of 1. In the above figure, note that 37.8% UPRO had a mean daily gain approximately 2x greater than SPY, and only a slightly greater standard deviation.

**Discussion**

Obviously, I'm very happy with the performance of the strategy since I started using it just before the start of 2016. It is important to understand, however, that much of the excellent performance in 2016 is due to BLV's tremendous growth. In general, I wouldn't expect the UPRO/BLV strategy to outperform SPY by more than a few percentage points a year.

It's always nice to see strong performance of a strategy prospectively. I have a lot of faith in this particular strategy because of its backtested performance, its performance so far in 2016, and the fact that it makes intuitive sense (use a leveraged ETF to free up assets to put somewhere else).

Of course, it's always possible for bond funds to experience a downturn, and even generate negative alpha for a period of time. When this happens, the extra assets allocated to BLV will generate losses rather than additional gains, and the UPRO/BLV strategy will underperform SPY. But of course we would expect BLV to have positive alpha more often than not.

**Disclosure:** I am/we are long BLV, UPRO.

I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.

**Additional disclosure: **The author used Yahoo! Finance to obtain historical stock prices and used R to analyze the data and generate figures. Any opinion, findings, and conclusions or recommendations expressed in this material are those of the author and do not necessarily reflect the views of the National Science Foundation.