I am trying to increase my understanding of the HF world lingo. Have heard this term before "Beta Neutral, FActor Constrainted Portfolio" many times with respect to how pods construct their portfolios. I am a former IB analyst / PE associate, so I am not a total newbie to markets, but can someone please tell me in layman terms what do such portfolios entail and how does one specifically go about constructing them? In other words, how do you identify names that fit these parameters and what do you have to do to make them beta neutral / factor constrained?
Secondly, what does market neutral portfolios mean?
Thanks!!!
You sure you are a former IB analyst / PE associate? You don't know how to use Google or ChatGPT?
x
Yeah, good luck with your trade reconciliation CAREER.
The only thing I'm reconciling is how badly I must have given to your mother for her to have given birth to you
Back in the ancient days (talking abt b4 the 50s or 60s), our understanding of markets was very limited. The only segregation of a portfolio's returns was only crudely separated into 1) market exposure, or beta, 2) manager skill, or alpha. If the S&P returned 5% and your fund returned 8%, people crudely assumed your beta was 5% and your alpha was 3%. Alfred Winslow Jones did this.
So obviously, any PM who could earn more than the market return, we just thought they were skilled. Buffett's monster returns - we thought all the excess return was alpha.
This was 'confirmed' by Sharpe with his creation of beta and CAPM. If you earned more than the market, that was probs alpha.
Fama and French came along in the 90s and discovered the value and size factors. That alerted the world to risk factors other than just the singular market beta factor. Over time, economists discovered more (risk) factors. These risk factors mean that the PM just took a risk and was rewarded with the risk premium. It wasn't skill, just luck by taking risk - what if the upside didn't materialize and his risk wasn't rewarded? Survivorship bias at play here.
With all these factors, people realized we could explain many monster returns including Buffett's as just beta. E.g., Buffett's returns are largely explained by him going long the value, quality, and low vol factors (look up the AQR paper 'Buffett's Alpha'). After controlling for these beta factors, Buffett's alpha is statistically insignificant. (This doesn't discredit Buffett's ingenuity at all - he intuitively discovered factor investing. He was a true genius pioneer)
Same goes for many LO/ LS managers - many of their returns are largely driven by risk factors
Going back to your question, a beta neutral/ factor constrained portfolio is the style practiced by the pod shops. In theory, their returns are all alpha. They don't take directional bets on risk or style factors like Buffett or Klarman - their trades are relative value or driven purely by security selection or market timing.
Therefore, since their returns are theoretically driven by all skill, beta neutral portfolios are uncorrelated with risk factors, while market neutral portfolios are uncorrelated with the market. (I view market neutral as just a subset of beta neutral, in my definition) E.g., a beta neutral portfolio won't go to shit because the value factor goes dead, and a market neutral portfolio won't go to shit because the market goes dead. That's why MMs thrived while value investors got hammered over the past decade - value investors were just long value while MMs were neutral.
These portfolios are theoretically very appealing to LPs because they deliver uncorrelated returns or alpha which is what LPs pay for. LPs don't need to pay a value investor to just deliver them value returns cuz they can just hold a value ETF, for example.
Thanks!
CAPM doesn't teach that if you earn more than the market then it's alpha. You can have beta greater than/less than 1.
What’s with the influx of googleable questions on this forum
market neutral means your strategy is uncorrelated to the market I.E 50% of your net is long and 50% is short within a sector. If tech is up 70% because of beta/factor, your names will all be up because of this. But the value of your longs being up is eroded by your shorts being up.
Actually your question is a legit one, and I’ve seen much worse questions in this forum and didn’t receive as much snark as yours did.
In factor land, each stock has its own “factor value”. Say stock A has x exposure to the growth factor, and stock B has y exposure to growth factor. If you were to construct the a portfolio that is neutral to the growth factor, you just take a linear combinations of weights that makes the total portfolio growth exposure zero. In the above example this is equivalent to solving 2 simultaneous equations which is trivial.
The above example is trivial since you only have 2 stocks, and you don’t really care about anything else except for factor neutrality. In the real world, you probably want to maximise your utility function (returns/outperformance), given some constraints (tracking error/vol/liquidity/risk limits/factor constraints). The problem gets complicated very quickly. To deal with all these dimensions, most folks use portfolio optimiser to find weights that match their criteria. Read about convex optimisation if you want to know more.
Then you might ask, how do we estimate each stock’s factor value? The short answer is by using a factor model, and it depends on whether the factor is a macroeconomic factor (inflation/sensitivity to rates) or a fundamental factor (P/E, EPS etc). I won’t get into that since that’s quite intricate. There are lots of resources out there on this.
Exactly. Idk what those people flaming OP were on abt, especially the 1st dude. It's a topic that's not easily understandable by Google or Investopedia.
There are questions here comparing prestige or SA modelling tests and no feathers get ruffled
Thanks, this was very helpful. A few quick follow ups for you -
Is this the analysts and the PM's job to construct mathematical models that identify a stock's exposure to a select factor? Or is this outsourced to a different team that does it and maintains the integrity of data? Don't believe this could be a BBG program, it has to be proprietary. If proprietary, is this maintained in excel or python or some internal CRM type database
Lastly, from a practical POV and the perspective of analysts who're covering the stocks that fit these parameters (beta neutral, factor constraint), does that mean that I am from the start given a set of names to cover, and don't need to go hunting for ideas/names only to find them rejected because of how they behave/correlate with certain factors?
This world is new to me, so yeah my apologies if this is a noob question
Bloomberg has publicly available factor models, as do Barra and many other providers. This is fairly standard and there's no rocket science to it... Citadel or MLP might have their own model, but it won't be meaningfully different than Bbg's.
There’s nothing that sophisticated about monitoring factor exposures. Yes, it’s arguably sophisticated enough to employ a small team of folks who majored in something more quantitatively rigorous than sociology. And yes, some firms may even be dumb enough to think that their monitoring mechanism is proprietary. But in reality, it’s simple stuff and the jargon is over-used so that people who do that work can feel smarter than they are.
Whatever work you did in IB and PE is considerably harder. So if you’re studying this stuff for an interview, definitely assume the simpler explanation is the right one.
Lol you can’t be more wrong on this. It is sophisticated, and at least an order of magnitude harder than IB/PE. But ofc wouldn’t expect PE guys to think otherwise.
The simple thing is the relatively stupid thing, that’s what people miss. Yes, many investors just use the off the shelf Bloomberg stuff. The question is how much of the value that gives you (I.e., is it 80% of the value or 99%?). The larger your firm gets and the more you are trying to squeeze value out of every corner the more this matters. The reason for all this factor neutral/market neutral/etc is risk management, especially as you are running a lot of leverage. If you screw that up, you can get demolished.
As for complexity, how much of a stock is driven by value or momentum or growth? It isn’t simple, and I’ve seen the models on Bloomberg, etc and they can be pretty wrong.
~chortling~
yeah, alphabetizing logos and doing DCFs are the true signals of intellectual horsepower
This thread is amazing
This question is a legit one, and portfolio construction is a huge part of the job. Some of the responses on here are great and some really unnecessary (even your own).
As mentioned Fama’s work is the king of this research area. Another area that is helpful is “Ray Dalio’s holy grail” he explains why these firms operate the way they do and the strength behind it.
Neque laboriosam et ut ullam voluptatibus. Nemo sit dolor ut accusantium et. Enim inventore reprehenderit temporibus enim. Qui quod qui quas.
Quos qui tempore est assumenda omnis asperiores. Illum earum reiciendis molestias corporis. Numquam et qui facere mollitia quibusdam beatae.
See All Comments - 100% Free
WSO depends on everyone being able to pitch in when they know something. Unlock with your email and get bonus: 6 financial modeling lessons free ($199 value)
or Unlock with your social account...