Netflix for Quant Funds –A new way to search for what you actually want

Netflix for Quant Funds –A new way to search for what you actually want

Let’s say you want to look for a new machine learning fund. You do not have a particular fund name in mind you just want to know who is launching, because maybe you can get a discount on fees or managed account. A couple of ways you can do it:

1. Google

2. Call up your trusty cap intro contact

3. Search on a database

You are feeling a bit lazy to search on Google and you don’t want to make small talk with cap intro, so you go for option C. Now here’s the problem. There is no way to search for what you want with the traditional filters:

The first apparent problem is that is no “negative” for track record meaning pre-launches cannot be found. If I select on track records less than 3 months, I do not get new fund launches or even new share classes, instead I get funds which started reporting to the database for less than that period of time.

Ok I will not give up easily. Let’s put in the word “new” and click on the magnifying glass. I get funds which have the word “new” in their fund names, eg. Guangdong New Value, Shenzhen New Thinking and Optima Fund Management (New York).

Clearly, this is not working for me. So we decided to build a new research database where we search for key words not just in the fund names, but also in public news, fund presentations and monthly letters. It’s just like Netflix. If you liked Legally Blonde, Netflix might suggest other movies with Reese Witherspoon, but will probably not ask you to watch Boston Legal.

Look at what you get in a traditional database when you look for peers to Bridgewater:

I don’t know about other people but I probably am expecting funds other than Bridgewater. MLM Macro has been around for about the same number of years as Bridgewater but its correlation to Bridgewater is about 0.1. Just because I like Legally Blonde released in 2001 doesn’t mean I like Training Day also released the same year. Although Black Hawk Down was pretty good.

Most people select movies because it has certain repeatable characteristics. Maybe it has the same actor, or it is the same genre. I don’t know of many movie watchers who say I want to watch the top grossing movies in a certain year. Similarly, it is somewhat unnatural for an allocator to say let’s allocate to the fund with the highest returns this year, which if you look at 2017 will be one of those crypto funds for sure. But funds are usually searched based on returns.

Now we have HedgeQuery! Here’s what we get for pre-launch Machine Learning funds:

1. Inferent Capital

2. Duality Capital

3. Bayforest Technologies

Much better.