How many ideas do you generate per week/month?

There's always a misconception about how funds run so I'm going to add a little more color here.  It's not about speed and percentage of ideas that get put on.  Both of these are functions of experience and track record of the analyst.  What is important though, is that the research process is correct from the very beginning for an up and coming analyst and eventually both speed and accuracy will increase with time.  

I'm a Senior Analyst at a low-net, single manager L/S shop, sector specialist.  

For me, about one new fleshed out idea every 1.5 - 3 weeks ready for internal discussion.  If it's earnings season it could be 3 weeks.  Doesn't matter if the name is put on or not, it's about doing the work right.  If we think the research is right then it will eventually be put on and it will have a chance to make money and at worst, come out flat or losing a little.  More often than not, one single, deep researched idea results in a pair-trade and deeper insight into a particular market landscape, yielding 2 actionable items on the portfolio level and 1 - 2 new idea leads.  Speed and accuracy increase with experience, mainly due to pattern recognition.  Obviously the faster and more accurate/usable a model/idea the better, but sometimes the process of building out the granular parts of a model for one company will be longer than the other.  This could vary between a few days to an extra two weeks.  Can't sacrifice quality for speed. 

For analysts, the point is to cover a sector to the best of one's ability and model out each name, then pick the top winners to long and losers to short.  Prioritization here is important and being able to generate ideas that have the best risk/reward, mispricing, and eventual alpha within a predefined time horizon comes with experience.  Generally, Junior analysts or analysts coming straight out of IB are not given the responsibility to generate ideas but assist in the research of ideas generated by Sr. analysts or PM's.  

Each stock/idea will have it's fully built model with a short and concise memo.  The investment memos can be 3 bullet points or just 2-3 sentences.  Keep it short and concise because we go through the model line-by-line, assumption by assumption, and let the excel analysis do the talking.  Daily conversation within the team is regarding newsflow for current positions and debate why we may have to prioritize x name versus y to carry out deep research on as well as updates on on-going research.  This process repeats itself until we fully exhaust our opportunity set.  We think a consistent idea velocity and accuracy is important here and this is where some HF's may vary from Long-only shops.  Being able to short individual names essentially doubles the opportunity set and workload.  An idea that was passed on by a LO may very well be a short for the HF.  This is even more so for low-net & market neutral L/S shops.

Most of the ideas will eventually be in the book given the stock reaches our predefined price for optimal margin of safety to go long or go short.  No neutral ratings, it's either a buy or a sell.  We let the market come to us for entry points.  Some names may take weeks or months to reach our predefined risk/reward profile and we monitor this very closely, especially when putting on shorts.  If it never reaches our target entry price for an attractive risk/reward, that's fine as we want to minimize our chances of getting blown-out of a position.  In this case, there's a review on what we may have missed in our research process and then we move on to the next name.  This is better than coming up with a stock's fair value that is inherently wrong because of poor/careless research due to pressure to generate performance.  Not putting on an idea that will generate PnL for the quarter or semi-annually (usually no longer) is fine as long as the research is done thoughtfully and correctly as we look at 2-3 year time frames for our longs and 6-months to 1 year for our shorts.  Having no ideas that make money within a time of more than half a year to a year means there's something wrong with the research process of a particular analyst and that we have to review as a team why this happened and what needs to be done to improve.  Consistently wrong views are less probable because each idea's model and core assumptions are reviewed in detail and debated amongst the team.  If an internal consensus can not be reached on what a fair value for a stock is due to a factor that we just can't figure out, then we don't put on the idea.  We try to minimize this occurrence during the idea generation process as not having an actionable item after research is essentially an opportunity cost.  

There is a list of already well-researched ideas (targets) for us to pull the trigger when the stock hits our entry prices, so a consistent flow of well-researched ideas is preferable for optimal use of time, but again, can not sacrifice quality.  Consistent does not mean fast.  A proper fund will already know the level of an analyst and delegate the correct level of responsibility to give along with the correct expectations.  This is why speed isn't really an issue and whereby accuracy is a must.  Though we can't fall behind on building out new names, or else we're essentially losing out on new opportunities.  Steady, consistent, and accurate work.  Overall fund idea velocity will essentially function on the size and experience of the investment team and in reality, it's about being accurate and fast, not one or the other.  The faster we churn out accurate, properly researched ideas, the more opportunities we have to generate higher returns.  Early bird gets the worm, so the saying goes.