Direxion XXX ETF's - Short and Long portfolios BOTH down.

[homerberkowitz]

I decided to research the new and risky Direxion 3X Leveraged ETF's by creating two portfolios, LONG and SHORT to see which one worked the best over the period of a few months.

Each portfolio started with $30,000, and each of the four funds in each portfolio was allocated 25%

As of today, the long portfolio is down 28% and the short portfolio is down 20%.

Direxion 3X short portfolio:
BGZ Current Price: $76.92 Price Paid:$ 86.74 Down 11.44%
TZA Current Price: $65.88 Price Paid:$ 93.76 Down 29.74%
ERY Current Price: $41.05 Price Paid:$ 49.33 Down 16.99%
FAZ Current Price: $66.65 Price Paid:$ 80.25 Down 17.36%
Total Loss: 20%

Direxion 3X long portfolio:
BGU Current Price: $23.45 Price Paid:$ 27.70 Down 15.06%
TNA Current Price: $20.03 Price Paid:$ 22.48 Down 10.40%
ERX Current Price: $27.91 Price Paid:$ 34.50 Down 18.60%
FAS Current Price: $5.48 Price Paid:$ 17.85 Down 69.12%
Total Loss: 28%

What gives. These portfolios were created at the same exact time, with the same amount allocated to each long and short fund. How could BOTH be down over 20%? This is a perfect example of a lose-lose situation.

Comments please.

Homer Berkowitz

[mgerstein]

This has been a much discussed and highly controversial topic with leveraged ETFs ever since ProShares introduced double-movement. Direxion, by upping the leverage to triple, adds a "more so" element to the whole debate.

Leveraged ETFs are designed to meet their targets on a daily basis. And the Direxion ETFs have been more or less doing that.

I looked at the 61 available daily price changes for BGZ, the large cap 3X short ETF and BGU, the large cap 3X short ETF.

I used the S&P as an approximate benchmark. Strictly speaking, I should be using the custom large cap index created by Dimension for use withthe ETFs, but I don't have the numbers and for now, I'm willing to assume the S&P 500 is in the same ballpark as their large-cap index.

The short ETF moved down every day the S&P moved up, and it moved up every day the S&P moved down, exactly as it's supposed to do. The long ETF had a couple of mis-steps; it got the direction correct 59 out of 61 times.

And it looks like the ETFs came close to tripling the daily movements. The average daily "error" (absolute value) for the long ETF was 0.43%, and the average for the long ETF was 0.53%. Those seem big, but remember, I'm fudging by using the S&P 500. (Note: the absolute values of the daily moves avergaged a little more than 6%.) The correct error calculation would have to be based on a comparison with Direxion's large cap index.

But as with any levraged ETFs, things change when holding periods stretch beyond a day. Even with a week, you can't assume the same pattern as you would for a day because you don't know the sequence of daily moves.

Assume, over the course of five trading day, the benchmark moves as follows.

Benchmark starts at 100
Day 1: +1.5%, benchmark = 101.50
Day 2: +2%, benchmark = 103.53
Day 3: -1.7%, benchmark = 101.77
Day 2: +3.5%, benchmark = 105.32
Day 2: -5%, benchmark = 100.07

Assuming a 3X long ETF triples, each daily move, here's how the week would shape up:

Benchmark starts at 100
Day 1: +4.5%, benchmark = 104.50
Day 2: +6%, benchmark = 110.77
Day 3: -5.1%, benchmark = 105.12
Day 4: +10.5%, benchmark = 116.16
Day 5: -15%, benchmark = 98.74

So there we have it. Over the course of a week, the benchmark is up a litgtle under 1%, but the triple long ETF is down 1.26%, even though each daily movement was exactly on target.

Looking at the 57 available consecutive-5-day periods for BGZ and BGU, we find that the short ETF matched overall direction 37 times, while the long ETF matched 36 times. If you stretch the holding period further, you expose yourself to more compounded mismatch, even if single-day moves are correct.

I'm among those who like and have successfully used ProShares ultras, recognizing the impact of compounding the doubles over multiple periods and adjusting my expectations and strategies accordingly. I still haven't decided if I want to use Direxion, given that "more so" quality.