# What is smart beta, alternative index, factor investing?

What is smart beta, alternative index, factor investing? Are they basically the same thing? Construct a benchmark index using schema other than market cap?

In recent years there has been much attention given to defining indexes other than market-cap based indices. While market-cap based indices approximate the theoretical Market Portfolio enshrined in textbooks, some people believe we could do better than that. One popular idea is that "market indexes overweight the most overvalued stocks", though this is controversial. Thus a desire for Alternative Indexes. (Note: an alternative index will have higher trading costs than a market index, so it is not clear which is better).

There are several kinds of Alternative Indexes:

Some indexes overweight certain factors (such as Momentum, Value, Small Cap, Low Vol, etc.) separately or in combination

Some indexes called Fundamental indexes are based on objective company characteristics such as Sales or Aggregate dividends, that don't take market price into account.

A clever marketing name Smart Beta has become popular. This is somewhat disliked by academics such as William Sharpe, who find it misleading or outright silly (and I agree). These funds are generally based on multiple factors.

Investment firms that have created products along these lines include: Research Affiliates, Invesco, iShares, TOBAM, Wisdom Tree, AQR, GSAM.

Finally it should be noted that Factor Investing is an older and more general idea than this. It goes back to the APT (Arbitrage Pricing Theory) of Ross (1976) and has been used in many different applications, not just building better indexes.

• I looked at this MSCI foundations of factor investing, they sort of treat Factor Investing and alternative index the same thing? I think what you said is factor model which goes back to long time ago. – JOHN Feb 10 '16 at 15:56
• Also, "(Note: an alternative index will have higher trading costs than a market index, so it is not clear which is better)." Can you sort of explain? Say, i have a index based on dividend, and my stock forecast also weight more on dividend stocks. Won't this result in a better alignment and hence less trade? – JOHN Feb 10 '16 at 15:58

The term "smart beta" is loaded and ambiguous. It means different things to different people. Some people manage products that they would argue are not smart beta while the rest of the industry vehemently disagrees. I've gone the route of defining what it means to me.

Smart Beta
Short Version
Commoditized Factor Investing

Longer Version
Investment strategies tend to follow a life cycle that begins with

Alpha - Brand new, fairly effective, and proprietary

and ends with

Beta - Generic asset class investing

Smart Beta is somewhere in between where we are mostly familiar with the details of factor investing and want to access investment vehicles more cheaply. As time marches on, various factor investing strategies will become more and more commoditized.

Those who claim that their factor investment product is NOT smart beta are those who believe that their product is differentiated from the rest (not a commodity) and that it takes skill to implement. Or, it's bad for business to admit that you can go around the corner and pay 9 basis points for the same strategy.

Response to Comment

Can you explain what does it mean: Investment strategies tend to follow a life cycle that begins with Alpha and ends with Beta?

This is my opinion and how I describe this process. When the first person to decided to invest more in equities because it had a higher return on average than fixed income, they didn't call it Beta. That person thought they had a good strategy and probably kept it to themselves. Proprietary return strategies are what I call Alpha. As that strategy becomes more well known and more accessible, it transitions towards Beta. Our quintessential definitions of Beta are now Fixed Income, Equity, Cash, Commodities, Currency. Smart Beta is somewhere in between. Momentum is becoming more accessible because we discuss it a lot more. We understand it better. We accept fairly standard portfolio construction around it. Its accessibility is no longer limited to high net worth individuals institutions. Index providers are producing easily maintainable portfolios that exhibit characteristics that are consistent with what we've called Momentum. The same thing is happening for Value and other factors as well.

The less secret and proprietary they are, the more Beta like they are. That's what I mean.

• Can you explain what does it mean: Investment strategies tend to follow a life cycle that begins with Alpha and ends with Beta? – JOHN Jan 4 '17 at 17:58
• @JOHN I've updated my post. – piRSquared Jan 4 '17 at 18:58

As @piRSquared has pointed out, smart beta can be an ambiguous term which is fairly loosely defined. Cliff Asness wrote a paper defining smart beta as

To be considered Smart Beta, we believe that these factors must also be simple and transparent. However, they don’t have to be the same for all managers or products. One can, and many do, argue that their particular version of simple value, low risk or any other tilt is better

He also has a more in depth discussion on the blurred line between alternative beta and alpha here.

I tend to think about smart beta (which I use synonomously with alternative risk premia) as an investment which offers positive expected returns above cash and which requires little to no "timing", a term I will make clear below. This definition would include things like equity beta, global duration, and currency carry.

I use timing in this context to mean the sensitivity of the investment strategy's performance with respect to being executed on a specific schedule. While there are many ways to quantify this, one method is to look at the correlation of the strategy returns with lagged implementations of the same strategy. i.e.

$$Corr(X(0), X(L))$$

where $X(L)$ is a time series of the strategy's returns with the value at time $t$ is given by

$$X(L)_t = \sum_{i=1}^N r^{i}_{t}h^{i}_{t-L}$$

where $N$ is the number of assets, $r^{i}_{t}$ is the return of the $i^{th}$ asset at time $t$ and $h^{i}_{t}$ is the holdings of the $i^{th}$ asset at time $t$.

The most extreme example of this would be a constant allocation to stock beta. In this case delaying rebalancing would have no effect on the strategy, since we are always long a constant amount.

At the other extreme, if you create a strategy that has a 1 day look ahead bias which goes long 1 unit when prices go up and short 1 unit when prices go down then you have no correlation at different lags, confirming the idea that this is a pure alpha strategy.

Somewhere in between you would have something like currency carry, which I would consider a risk premia style strategy.

• thanks for you answer and your link to that post. If I understand correctly, the post is saying that a factor has two attribution, risk compensation and a alpha premium. The alpha premium could disappear after more people is using the strategy, because participants create flows that increase risk, and when you take the same risk as before, you get less returns. But I get confused with your post especially the $cor(x, x_t)$ and the plots. Could you future explain? Are you trying to separate the return of true alpha and the return come from risk? – JOHN Jun 6 '17 at 21:34