How asset price affects your alpha

[ironchef]

So, does that means higher beta assets (like if I levered SPY) have lower risk adjusted returns even though in a bull market I get higher absolute returns?

The abstract is the only part of this paper that is comprehensible to me. The body is simply too hard to understand, especially for one not familiar with the terminologies. I need to go back to school to learn some finance.

I think I answered my own question. The study differentiate between higher beta (higher risk, more risky) assets vs levered asset.

But what is the real difference? Take the example of 40/60 Bond/SPY and 10/90 Bond/SPY, I can always find the correct levered/ratio of one to be equivalent to the other in term of producing the same return and Sharpe?

 

[mushinseeker]

I find Taleb nearly insufferable, so I hesitate to wade into his book to try to figure out what he is getting at. It would be easier to test the concept first, then if the idea seems to work on backtest, it might be worthwhile to try to figure out why it might work.

The formula for "alpha fair value," from the table you posted, appears to be alpha-fair-value = Vol^2 * 14.0265. So what you can do is, for every Monday-after-third-Friday in the past, sort the nearest-to-atm calls on all active names for the upcoming monthly expiry by alpha / alpha-fair-value. Simulate a buy at close-mid on the lowest 5% and a sell at close-mid on the highest 5%. Hold until expiry. Repeat monthly to the present, and see if the simulated returns are significantly different from zero.

If it works, start a hedge fund. Name it Empirica Redux.

Kevin. that is a great idea for a backtest. I would also add a tweak..try to find a subset in that universe that did not move after 3 weeks .. put in in a classifier algo and if Talebs theory holds, there should be 2 distinct groups- 1 that lost $x and the other that lost 2-33 times as much because the gamma rent was high?

 

[Magic]

I think I answered my own question. The study differentiate between higher beta (higher risk, more risky) assets vs levered asset.

But what is the real difference? Take the example of 40/60 Bond/SPY and 10/90 Bond/SPY, I can always find the correct levered/ratio of one to be equivalent to the other in term of producing the same return and Sharpe?

It seems like you've had derivatives of this discussion several times now. Not a criticism mind you; but I'm trying to think of another way to explain the underlying concepts.

You can't just lever a 40/60 into the same Sharpe as a 10/90. Because your leverage will boost the returns and the volatility in an equal manner.. Think of it this way, you can either solve for target risk or target return.

On the risk side, a 40/60 levered into the same SD or max DD as a 10/90 will produce higher returns. Or if you're solving for target return; the levered 40/60 will produce the same return as 10/90 but will do so with less volatility. There's a myriad of theories on why less volatile instruments post higher risk adj. returns as has been mentioned.. as long as your leverage comes at the right cost and it doesn't get taken away from you when you need it most you can outperform by levering more efficient portfolios vs. just jacking the beta of your composition.

 

[ironchef]

It seems like you've had derivatives of this discussion several times now. Not a criticism mind you; but I'm trying to think of another way to explain the underlying concepts.

You can't just lever a 40/60 into the same Sharpe as a 10/90. Because your leverage will boost the returns and the volatility in an equal manner.. Think of it this way, you can either solve for target risk or target return.

On the risk side, a 40/60 levered into the same SD or max DD as a 10/90 will produce higher returns. Or if you're solving for target return; the levered 40/60 will produce the same return as 10/90 but will do so with less volatility. There's a myriad of theories on why less volatile instruments post higher risk adj. returns as has been mentioned.. as long as your leverage comes at the right cost and it doesn't get taken away from you when you need it most you can outperform by levering more efficient portfolios vs. just jacking the beta of your composition.

Thank you.

I am just not very smart. If I really understand the concept and understand options, I won't be here asking questions.

So, like the 40/60 and 10/90 example, if I can get the same return with a levered 40/60 but less volatility, why aren't that be arbitraged away? I need to run the excel and see that myself.

Hope I am not wasting too much of your time. I very interest in this discussion because I am quite profitable but really don't understand why. I used long/short options basically as levered underlying.

 

[Magic]

Thank you.

I am just not very smart. If I really understand the concept and understand options, I won't be here asking questions.

So, like the 40/60 and 10/90 example, if I can get the same return with a levered 40/60 but less volatility, why aren't that be arbitraged away? I need to run the excel and see that myself.

Hope I am not wasting too much of your time. I very interest in this discussion because I am quite profitable but really don't understand why. I used long/short options basically as levered underlying.

It's difficult to give a single answer on why the disparity exists. My 2c is that the spread in risk-adj returns is already generally bounded around the fair value for it. An exact match wouldn't be fitting since having some "leverage" embedded in an instrument instead of managing a short dollar position yourself along with the variable costs and potential liabilities should cost some form of premium. Not to mention many other cited effects such as large pools of capital under leverage constraints due to regulation or law, psychological bias generating demand for lottery-esque payouts, unsophisticated participants without a clear conception of risk-adj. returns simply seeking higher yields, etc.

Also important to note that when we move away from high beta vs. low beta to portfolio analysis, part of the increased risk-adj. returns are coming from the benefit of diversification. When holding multiple asset classes the total risk doesn't necessarily dissipate, it just becomes more asynchronous and thus smooths the equity curve. In that case trying to set the risk-adj. returns of a single bucket to a diversified portfolio is like comparing apples to oranges.

 

[ironchef]

You can always hedge your SPY long with a portfolio of short SPHB and unleveraged long SPLB. That way you have a negative 1.0 beta hedge on on your positive 1.0 beta SPY with a significantly positive expectation (hedges normally have zero or slightly negative expectation -- i.e. it costs money for insurance), Also your hedge is probably, since low beta tend to have higher dividend yields, that rare thing: retail-self-financing (zero-weight). So you'd be protected on the downside, your hedge wouldn't cost anything to carry, and it would even make you money (like having your car-insurance company send you a monthly check instead of you paying them).

N.B. I haven't actually run the numbers on this, so caveat emptor. It's based on something low-beta proselytizer Eric Falkenstein published years ago, before he went off the rails.

Thank you. This is a good exercise/homework for me to work on.

I googled SPHB, it is a high beta etf from SPY stocks (the highest 100 beta stocks) but SPLB is long term corporate bond etf. Are we talking apple to apple?

 

[srinir]

Thank you. This is a good exercise/homework for me to work on.

I googled SPHB, it is a high beta etf from SPY stocks (the highest 100 beta stocks) but SPLB is long term corporate bond etf. Are we talking apple to apple?

There is no specific low beta etf. You can substitute SPLV, which is minimum vol. etf which has half the beta of SPHB. It is screened from the same pool (SP500) stocks, but screened for slightly different factor. But for this exercise that is good enough

 

[Magic]

@Kevin Schmit ; @ironchef

My first thought was that a Bet-Against-Beta component vs. long SPY isn't quite what I'd consider a classical hedge; it's more that we're swapping away market risk and holding the risk of the spread's oscillations, right? And those can still get pretty wide at times. Perhaps more efficient to leave some size/beta on the table from the SPY position, treat the BAB as a separate strategy, and just size accordingly in order to realize some diversification benefits?

Or we could apply Occam's razor, and just short high beta directly against SPY holdings itself to reduce exposure, and theoretically get a bargain with what we pay on the hedge vs. the amount of market rise we're offloading. And perhaps just bypassing SPY entirely and substituting in oversized low beta long if we believe it's truly superior to the SPY in risk-adj. returns.

Would be nice if the data stretched back a little further with these to get a better concept of how the relationships behave in different conditions.

 

[Kevin Schmit]

I googled SPHB, it is a high beta etf from SPY stocks (the highest 100 beta stocks) but SPLB is long term corporate bond etf. Are we talking apple to apple?

My mistake. I meant SPLV, not SPLB.

 

[Adam777]

Since Beta is risk/volatility of a stock compared to the rest of the market, wouldn’t you sell (short straddle/fly) low beta, high IV stocks; and buy (long straddle) high beta, low IV stocks?

Also I guess low beta would be large caps and high beta would be growth stocks ...