stock research / stock market / algotrading / backtesting / build in public
ONBPO Market Cap: Google Says $9 Billion. The Real Number Is $115 Million.
Google Finance and Robinhood both list ONBPO at a market cap of roughly $9.4 to $9.6 billion. ONBPO is a depositary preferred share of Old National Bancorp, a regional bank. It pays a fixed dividend and trades around $25. Both sites put the whole bank's market cap on the preferred ticker. Multiply its own units by its price (about 4.9 million units at $23.50) and you get about $115 million.
The wrong figure matters as soon as you build on it. A screen for bank stocks with a market cap between $5 and $15 billion would include this ticker, and a screen for banks with the lowest price to book ratio ranks it near the top, which is how I found the mistake. This is where most of my time on NexusTrade goes: making sure the backtesting system and the algorithmic trading platform run on accurate data.
Transcript
0:00Here's a concrete example of why building an algorithmic trading platform is genuinely hard And to show this I'm
0:05going to show it with a ticker that both Robinhood and Google got wrong in terms of its fundamentals
0:12That ticker is O N B P O This ticker is the Old National Bancorp stock which is a
0:17regional bank Now if you Google their market cap Google will give you the wrong answer at
0:22$9.4 billion And if you look it up on Robinhood it will also give you
0:28the wrong answer at $9.4 billion And both of these
0:34answers are factually wrong and I can prove it The reason why they're wrong is this is a
0:40special type of share is a depositary share worth 1/40th of their Class C shares
0:46It pays a fixed dividend and it trades around $25 per share Now this market cap
0:52being wrong matters for many reasons Let's say you want to create a trading strategy based
0:57on bank stocks with a market cap between $5 and $15 billion This ticker is
1:03going to show up because the market cap is allegedly $9 billion Or let's say you want to create a
1:09trading strategy with banks with the lowest price to book ratio That's actually exactly how
1:15I found this mistake I saw that the P/B ratio was less than 1 which if you know anything about
1:21what that means doesn't make any actual sense So to fix this I have to fix its market cap I
1:26multiply its actual units by the ticker's price and I instead get
1:31$115 million a far cry from the $9 billion
1:37that these other websites list And this is a bug you wouldn't normally think of This isn't a bug that you're going to think about
1:43before you build the platform It's only a bug when you actually read your results Neither Google
1:49nor Robinhood has found this bug and they list the inaccurate price prominently
1:55on their website And that is why I spend so much time on my platform because I just I care
2:01too much about correctness I want you to be able to answer any financial question that you have
2:06and get a correct answer It's not like I'm spending all day long editing the event driven loop
2:12within the backtesting system It's not like I'm spending all my time on the performance engineering I spend a lot of time on performance engineering but
2:18that's not what I spend all my time on I spend my time making sure that the backtesting system and the algorithmic trading
2:24platform has accurate correct data Bye for now.
Join the conversation
Loading conversation…