Mar

28

What is the difference between the "smart-beta" index built around "momentum factor" (offered by Russell or some other index provider) and a trend-following CTA? It seems to me like a lot of smart repackaging (trend-following is now called momentum since more academic research is about momentum, trading is now asset allocation, etc.)….

Aside of fees, of course :-)

Ralph Vince writes: 

All trading systems can be represented as indexes. (even your simplest, go long here, flat there, short here, has aggregatge weightings of 1, 0,1 on the various positions — cash always 1-position weight).

All portfolio models, can be represented as indexes.

So

Trading Systems ~ Portfolio Models ~ Indexes (~ representing "equivalent to")

It's a matter of packaging.

And further building on this edifice scratched in the walls of my darkened cave….

And all long positions ~ short put + long call of same series.

And all of this occurs within the hyaline manifold of leverage space, which readily explains things that are often not so evident on the surface (such as why a short etf will have a long-term downward drift, as well as all leveraged ones, just as with any form of portfolio insurance) and on and on and on and on.)

Rocky Humbert writes: 

Ralph articulates this well.

I would add one point:

We know as a logical syllogism that the overall return from an entire market (to all participants) is the overall return from an entire market. Putting aside the mark to market paradox, if I were the sole market participant and I owned the entire market, then my return would be the intrinsic market return (i.e. cash flow, profits, dividends, etc). And if there were two market participants, then the intrinsic return is shared between those two participants. Again, mark to market paradox notwithstanding, just as it is impossible to squeeze blood from a stone, one cannot produce a total return that exceeds the intrinsic market return. The only question therefore is how to allocate the return — which, beyond the intrinsic return, resembles a zero sum game. (Some people here call the intrinsic return, "drift", but it is really dividends, retained profits, etc.)

An academically pure index must capture the entire market's intrinsic return. And it would do that by owning the entire market capitalization of that market. The S&P500 doesn't do this exactly — the index owners exercise nuance and discretion — and that process might give some opportunities to the smart-beta crowd. That the S&P is market cap weighted further gives rise to the mark to market paradox (i.e. the starting point when one purchases the entire market cap).

But if one could actually purchase a piece of the entire market on the day of the market's creation — and own it until the end of the world — that investment will produce a return that will, after taxes and expenses — beat holding any given smart beta strategy for the same duration. This is a purely theoretical point — because during any given holding period, some particular smart beta strategy will surely outperform. Again, it's a mark to market issue. So the goal is to figure out which one will and which one won't. (Assuming that this is possible!)

So yes Virginia — there is a pure index. But it is theoretical ideal.

anonymous writes: 

I used to trade and develop "smart beta" strategies back at the fund.

I don't think there is an established "this is smart beta and this is not," but I can tell you as to what people expect. The momentum strategies are a bit different than typical CTA trending strategies (not using crossovers for example). Instead momentum is tracked by other measures such as relative performance across sectors and going long/short the best/worst performing ones.

The implied idea of smart beta, which is not exclusive to CTAs, are the other benefits of using these strategies amongst others in a way that utilizes portfolio construction or a dynamic weighting strategy (like monthly rebalancing on vol).

The goal with smart beta is to not produce alpha outright, but to accept that the majority of alpha has been "sapped" and you are now using diversified strategies that have a known cyclical alpha. This is where you get into gray area but I differ the two by saying:

alpha means the sharpe significantly deteriorates as others discover the method smart beta means the sharpe has already significantly deteroriated, but because it has, you can more easily predict the regimes in which they work/don't work. For example, AQR's paper: value and momentum everywhere discusses the idea that momentum (continuation) and value strategies (mean-reversion), tend to have negative correlations, albeit both strategies have lower sharpes (0.4 to 0.7).

anonymous writes: 

Assume, all flows as dividends etc produce an intrinsic market return of zero over some point. Trader A loses 10. Trader B gains 9 (dont forget the vig). Ok till this syllogism its a zero sum game

Next,
A sold at 100, B bought at 100.
A stopped out at 105.
B stopped out at 95.
There must be a C or a CDE.. and so on and so forth.
Still sounds like zero sum. 

But if over a length of time some stay in the game, majority keep dropping out. Then it becomes a series of zero sum games.

Next, If A,B,C,D,E…. et al become very large numbers then its a zero sum game between those who stayed in the game up to the point the non participants came in. This also explains the Lobogolas.

Market therefore is a variable sum game. People vary their exposures, they vary even their presence for prolonged periods of time. No one rides an investment bus permanently the way sage does. Normal people buy stocks with an "intention" to sell at some point.

The drift in equities is explicable by a fact that it is the only asset class where reinvestment in growth occurs. For Indian equities I have had calculated in the past it mimicks the curve of (1+GDP growth)*(1+inflation). Perhaps true for other markets too. Given in the long run supernormal profits dont exist, its the ability of businesses to pass on inflation to their customers that produces the drift in their cashflows and thus stock prices. 


Comments

WordPress database error: [Table './dailyspeculations_com_@002d_dailywordpress/wp_comments' is marked as crashed and last (automatic?) repair failed]
SELECT * FROM wp_comments WHERE comment_post_ID = '10972' AND comment_approved = '1' ORDER BY comment_date

Name

Email

Website

Speak your mind

Archives

Resources & Links

Search