What are pod shops looking for, to generate alpha?
Are all HFs pod shops looking for the inflection in second derivative of new business metric (could be different in each sector but you get the jist)? Are just classic beats-and-raises enough for a stock to work, even when second derivative is slowing down or going negative?
Trying to understand how much of alpha is coming from just basic understanding of model mechanics vs channel checks vs positioning etc? Would appreciate any insight.
Pod shops focus on identifying idiosyncratic alpha by honing in on the key drivers that are most likely to move a stock. Here's a breakdown of what they typically look for and how they approach alpha generation:
Key Drivers and Inflection Points:
Beats-and-Raises vs. Second Derivative:
Alpha Generation Components:
Sector-Specific Nuances:
Velocity of Idea Generation:
In summary, pod shops generate alpha by combining a deep understanding of model mechanics, real-time insights from channel checks, and a keen awareness of market positioning. While beats-and-raises can still work, the broader context, including second derivative trends and market expectations, plays a significant role in determining the stock's reaction.
Sources: Credit - Pod Shop/MM vs. Distressed/Special Sits HF, Demystify the LT SM / tiger cub / "PE approach" vs. MMHF / pod shops?, Q&A: HF Analyst @ $5bn+ Fund - Breaking In and Transition to Risk-Taking Role, MIT vs. Princeton vs. Yale undergrad for quant hedge fund, How do generalists produce alpha?
It’s a bit more complicated than that.
This question reads as if you put a bunch of pod words together, I don't know what you're specifically asking here but I'll do my best.
There's no set "thing" to look for. It can be one of these, a mix, all or none. Really depends on the specific opportunity. You will find some PMs who think about stocks with a value or relVal tilt, and what they "look" for in an opportunity might not be any of these specifically, even if these are used somewhere in the process to find something else.
WSO and other providers (iykyk) are really good sources for the dissemination of pod-style trades and pitches that the median PM/analyst chasing EPS might be focused on...but that's not a sustainable way to run a $2bn+ book if your entire process is: incremental data point + crowding/positioning
For what its worth, I've noticed some of my recent interviewees are more "pod-literate" and can talk the same language as me and produce good tactical, short-term ideas but then that's all they produce. Better than the guy pitching terminal thematic with no idea on what's in the price and risk/reward, but not quite what I'm looking for if you can't identify quality businesses and compounders either.
Many ways to skin a cat in this biz.
If I’m pitching a long idea to a pod-style interview, how do you think I can differentiate? I don’t have access to the alt data sources or ability to do broad channel checks given I’m not in HF seat right now.
Would you say an actionable idea (ex software) may look like ‘+10% new logo growth is priced into the stock today, but through my DD, my view is +15% new logo growth is achievable next quarter, which should boost EPS by x%’
For context, I’m a former PE associate who is prepping for some HF interviews.
This is the job and no one is going to tell you how to do it
Would you say an actionable idea (ex software) may look like ‘+10% new logo growth is priced into the stock today, but through my DD, my view is +15% new logo growth is achievable next quarter, which should boost EPS by x%
Broadly yes
Most ideas we generate are bayesian and multi-layered.
On the Bayesian side you start with a prior and develop a posterior view upon new information either printed or generated through your research process, ultimately rolling up into an estimate revision.
On the layered approach, you consider whether this new bit of information is incremental to a shift to a new scenario.
E.g. you follow company A which is a mid-tier software vendor, squeezed in two directions by larger more enterprise friendly competitors and smaller more nimble/cheaper ones. The company launched an ambitious plan to get the company to $1bn in ARR in 5 years, last year. Consensus estimates and the market have not moved to reflect it, as the company will need to "prove it" and it'll take at least another 6 months for a demo.
Now you know what the setup and what the potential catalyst is. If the street is at $780mn for the terminal year, and the multiple remained stable since they announced the plan, what's the upside risk into the catalyst? Well, if the co can demonstrate it's on track, that might move estimates towards $1bn. If the company is seeing an improving dynamic today (improving NRR, billings, etc.) then there could be upside to the $1bn. If the TAM is accelerating due to some exogenous theme, that could also significantly lift estimates.
What should be clear in this example is that you have a potential trade in 6 months for a number that's still 3-4 years out. Once you understand there's a bet to make, as an analyst you'll try to really dial in on where you lean on the risks. You'll put together a few scenarios outlining the outcomes you think are possible. You'll price the stock on your base case (e.g. the multiple on your base case estimate) to evaluate richness/cheapness of the trade. If you think IRR looks good, you'll recommend the trade.
As an analyst you are doing this work on your entire coverage of 20-50 names. Not all will have clear catalysts or setups within the next 6-12 months to trade. Those can be great as pairs or to benchmark other co's against.
“You'll price the stock on your base case (e.g. the multiple on your base case estimate) to evaluate richness/cheapness of the trade.”
Can you elaborate what you mean here - are you trying to back out into the implied multiple based on a DCF of base case and comparing to peers? or is it
You slap some reasonable multiple based on comps on the base case and see if the juice is worth the squeeze?
Current EV / your estimate of the metric (sales, EBITDA, whatever) = your multiple today.
The delta between your multiple and consensus should be driven by your view of the metric (your estimate vs. street). E.g. 20% variance in sales = 20% variance in your EV/sales vs. consensus EV/sales.
Now you can estimate where the consensus multiple can go if the market begins to agree with you (consensus EV/sales * (1+ your delta in the multiple, e.g. 20%). So if the stock is currently trading at 10x then on your estimate it can trade up to 12x (if no estimate change from the street). This sort of analysis is very helpful to determine the magnitudes of moves. Remember -- the buy side mantra is "put a number on it" (quantify your view into an estimate).
c
"those can be great as pairs to benchmark other co's against."
When pair trading can you (or do you ever) use several co's as your 'pair' when trading?
E.g., 50% of trade in long position of company A, and 16% each in 3x different comparable shorts - to ensure outcome is even more likely to be idiosyncratic.
Occasionally. I do that when I want to “hedge” if I can’t find an alpha short.
Save
FYI there are bookmark and subscribe buttons at the top... but it also took me years to find them, because my ad blocker was hiding them
Eos voluptatem voluptatem non sit. Consectetur vero inventore suscipit enim minus ipsum deleniti exercitationem. Repudiandae sit odio eos illum.
Ut accusamus voluptas quis harum qui. Ullam aut reprehenderit optio facilis quisquam architecto est illo. Rerum cumque saepe dolorem. Consectetur dolorem sunt architecto esse ut minima quasi. Numquam voluptatem ut eaque.
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...