Showing posts with label track record. Show all posts
Showing posts with label track record. Show all posts

Wednesday, February 24, 2016

Meeting the school money investors

Private schools are non-profit, but they have a lot of money from undergrads paying tuition, and money from fellowships and grants. Also donations from the alumni. Quite a significant portion of this "profit" is not immediately spent on the needs of the school, but is instead invested. The scale of this investment is billions of dollars times how big the school is and how long has it been around. On the east coast, it is really funny to observe competition between Harvard and Yale - who can make most returns this calendar year, or in a three-year interval? Both have of order 50 billion.

Of course it's not a fair comparison - each school has different needs. There are other indicators they like to compare themselves to. The non-profit part allows them to pay no taxes. What's the most interesting is how the investment decisions are made, and what kind of choices are available on that level. Of course this is not as big as it gets, but it's big enough to open most doors for them. More below.

The people in charge are called something like "board of trustees". They are about 12 people who are old and experienced in investment. Some are professors, some have business background. A few consultants who do not have a say in the final decision making are also present on their meetings, e.g. a consultant on China market. I assume that a lot of paperwork is handled by secretaries or their equivalents. What this organization's daily job is can be described as a lot of paperwork, and some very qualitative analysis of the bigger picture. They do look on quantitative performance measures, but then they are mostly talking to make sense of it. School provides this organization with budget needs (amount to be subtracted from the fund's value at the end of calendar year). There don't seem to be any other constraints.

The actual buy and sell decision are made by money managers, who sign an agreement with the above people. These managers are different companies like hedge funds, and also independent venture capitalists etc. Each one specializes in one or two areas. The board of trustees does not tell them what to do, just asks them what they usually do, and then if their plans align with board of trustees big picture view, they are hired. Then board of trustees monitors their activity, and if it deviates from the expectations, or they just plain lose money. 10% every year get fired, and it leads to a lot of annoying paperwork for the board.

2b/100 = 20m$ per manager. This opens all the doors of hedge funds.


Saturday, September 19, 2015

Plan of Action

In order for us to get our idea off the ground, we have a clear plan of action:

1) Develop an algorithm for Quantiacs using Matlab and ML
2) Develop an algorithm for Quantopian using Python and ML
3) Create a library that quantifies the impacts of small traders on the market
4) Create a website with method of payments using Shopify
5) Publicize our company via blogs, news, conferences and social media
6) Resolve all legal aspect of registering and operating company in the US
7) Create our own broker company and contract
8) Create a Kickstarter campaign
9) Develop software to help investor make decisions about optimizing their investment in stocks and real estate

This project is part of LaunchTech:


Track record: Quantiacs algorithm performance

The Quantiacs trading competition is held every half a year. An algorithm trained on historical data is tested for 3 month. After the winner is determined, the algorithms' performance continues to be public on the Quantiacs website. Here is my current top algorithm's performance on the new data.
U signifies the upload date, after which no changes to the algorithm has been made. The plot shows how the returns of the algorithm survived through the market crash of August 2015.
Trading Days89
Perf.
2.16% (6.34% p.a.)
Volatility4.37%
Sharpe Ratio
1.445
Sortino Ratio
2.183