Leveraged ETF returns/rebalancing

[lime]

I understand that leveraged or reverse ETFs need daily rebalancing to replicate the promised reference return and achieve target exposures, resulting in a convexity/trend factor (some called volatility drag).

What I am not sure is exactly which return of the underlying is one targeting? Say for QQQ and its 3x TQQQ and 3x reverse SQQQ, what do daily return and exposure mean? Is it close to close or intraday return (close to open)?

I am trying to compare TQQQ/SQQQ returns to QQQ (x 3) in the attached spreadsheet (data from Yahoo Finance). There are always residuals and quite large (5-8 basis points) in both close-close and close-open returns, so I cannot replicate the 3x returns.

 

[2rosy]

What does the spec say? Usually levered etfs should multiply what the reference etf did on the day. They are basically short gamma and were front run at end of day years ago, maybe still are. Shorting both leveraged and inverse leveraged probably a good trade depending on costs

 

[S2007S]

You want to make free money. Just short the 3x bear etfs. Do backtesting on them ...everytime they reverse split they drop straight back down to new lows within a 12 to 24 month time frame. Now there will be a big run or 2 on any market pull back of 5 to 10% but since markets only go up these stocks usually only fall over the extended periods...and if you are caught in the red on these short positions, just cost average in and you'll eventually see the 3x etf fall past your cost average short per share price.

 

[lime]

What does the spec say? Usually levered etfs should multiply what the reference etf did on the day. They are basically short gamma and were front run at end of day years ago, maybe still are. Shorting both leveraged and inverse leveraged probably a good trade depending on costs

Not sure if your reference to gamma is accurate here as they are long convexity by trading with the market (trend following).

 

[2rosy]

Not sure if your reference to gamma is accurate here as they are long convexity by trading with the market (trend following).

If you think of the leverage factor as delta then if reference index goes up 1% a 2X etf needs to go up 2% which is where the rebalance comes in. So on huge up/ down days levered etfs exacerbate the move

 

[lime]

If you think of the leverage factor as delta then if reference index goes up 1% a 2X etf needs to go up 2% which is where the rebalance comes in. So on huge up/ down days levered etfs exacerbate the move

How I see it is shorting both TQQQ & SQQQ is like short straddle, so short gamma, so the long side must have long gamma.

The gamma you refer to here is the drag, that x in, symmetric move up and down, (1+x)(1-x) = 1 - x^2. I don't think calling it gamma is accurate, as consecutive moves in the single direction substantially benefits a Leveraged/Reversed ETF (that is long gamma). I don't see gamma as path-dependent.

 

[2rosy]

How I see it is shorting both TQQQ & SQQQ is like short straddle, so short gamma, so the long side must have long gamma.

The gamma you refer to here is the drag, that x in, symmetric move up and down, (1+x)(1-x) = 1 - x^2. I don't think calling it gamma is accurate, as consecutive moves in the single direction substantially benefits a Leveraged/Reversed ETF (that is long gamma). I don't see gamma as path-dependent.

In options when you're short gamma and underlying goes up you need to buy to rebalance your position. That's what I was getting at. Leveraged etfs need to buy on up days and sell on down days to rebalance

 

[lime]

In options when you're short gamma and underlying goes up you need to buy to rebalance your position. That's what I was getting at. Leveraged etfs need to buy on up days and sell on down days to rebalance

That is why I call them trend-following (buy higher, sell lower), but whether such rebalancing benefits or hurts depends on the trend. When your short option gamma, your delta hedging also forces you buy higher and sell lower, but the PL is negative in each rebalancing (path-independent), wheras the LETF rebalancing PL is path-dependent.

https://math.nyu.edu/~avellane/SIAMLETFS.pdf.pdf

 

[Slow Learning Elf]

Not sure if your reference to gamma is accurate here as they are long convexity by trading with the market (trend following).

They have no convexity. The decay comes from mean reversion of the underlying index (mostly), from slippage at the close and from market impact (for LETFs with less liquid underlying). Where do you think convexity comes from if the rules just trade delta one instruments?

 

[lime]

They have no convexity. The decay comes from mean reversion of the underlying index (mostly), from slippage at the close and from market impact (for LETFs with less liquid underlying). Where do you think convexity comes from if the rules just trade delta one instruments?

I am referring to the curvature: