idea velocity at fundamental or quant funds
Want to hear people with experience in the industry (both PM and analyst perspectives) on what expectations are like for a junior analyst or quant with respect to generating ideas.
How rapid should new ideas be generated such that a PM would be satisfied with idea generation velocity?
How quickly should new ideas be tested?
What are expectations for the number of ideas that contribute to the portfolio?
As an analyst or PM how would you answer the above so that you're comfortable in keeping your job, you have enough bargaining power for a raise, you are the MVP and can dictate terms or even ask for a slice of the pie.
Depends on the type of shop. At the multimanager platforms, your PMs expect new ideas pretty quickly given how short term the investment style is. I would say ~3-5 new investment ideas every month.
At some of the longer-term oriented shops like the deep value / distressed funds, I would say one idea every couple months. But it isn't that you just sit there and do nothing the whole time - you would go through 20 or so potential companies to invest in before finding one that fits what you are looking for.
For books that trade ~quarterly, I was also going to respond with ~1 idea per week out of an at least moderately experienced analyst
thanks for the reply. Out of the 3-5 ideas how many would you expect to be something that would be put into the book?
thanks for the reply. Out of the 3-5 ideas how many would you expect to be something that would be put into the book? Would this change for junior vs senior analyst? How many would be expected to be profitable?
Senior analyst probably getting almost all in, with some sized higher than others on PM discretion
Mid analyst (1-2 years in) you are killing it at 50% in, doing fine with 1/3 id say
First year or two, you might get an idea or two in, but you are more importantly learning your coverage
Profitability - just be in the green
For quant, it depends on the idea. A large change could take months of research (data gathering, cleaning, backtesting the idea alone, backtesting the idea into the whole model, etc.). A relatively simple change that might just be changing a few lines of code somewhere could take a few hours to a few days. Some shops are more careful in making changes than others. Better to be more patient with making changes to ensure that a small change doesn't break something somewhere. I've seen cases where a small code change didn't have any issues for 6 months until we encountered an issue with the dates in the column of a database at the beginning of the new year.
Thanks for the explanation. Seems like you are describing quite a collaborative team. Would you expect this to change if it was more of a multi? As a junior would you be expected to come up with a completely new strategy (vs twiddling at the edges) and how long should it take? What if it doesn't work? At what point do you cut your losses and move on?
You would never come up with a "completely new" strategy only enhancements to an existing one. Those enhancements might be large or small.
As for all of your other questions, it's dependent on the type of fund, investment horizon, etc. Quant funds that have a longer holding period could keep an enhancement that hasnt been working for years if they believe there's a good reason that the enhancement will work over an entire economic cycle. Valuation factors have been getting crushed for years but most quants are still using them because there is strong empirical evidence that in the long run they work.
To add to this, at a single manager quant fund you usually won't be designing an end-to-end strategy but refining components of their existing process. That can be alpha signals for individual securities, statistical risk models, trading cost models or various other things. Most established quant funds are pretty conservative in making changes to existing strategies, even when they are underperforming (especially the ones with a large AUM). In the larger or older funds, the production system is often run by a separate group from the one doing quant research, and the groups may or may not be siloed or have access to each others' codebases. In these cases, the quant writes some code but the other group checks it in and monitors it.
Very insightful. I've paraphrased the questions I asked deepLearning above: As a junior would you be expected to come up with a completely new strategy (vs twiddling at the edges) and how long should it take? What if it doesn't work? At what point do you cut your losses and move on? Once you are more experienced how often should you test ideas? And what percentage of these should become something implemented in the book?
Eos quasi non labore. Voluptas optio minus minima repellendus in pariatur modi. Esse et repellendus a odio suscipit quo.
Qui adipisci voluptates ut eveniet itaque et. Et ea quia fugit eum sequi. Est earum laboriosam porro eos aut itaque modi ipsam. Placeat sint dolore praesentium voluptas accusamus ea et. Quis nostrum impedit quia eum. Dolores blanditiis cupiditate temporibus est optio sunt magnam.
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...