Handwaving about Optimizing Income Statements with Data Science

Buy the rumor, sell the news

Someone asked me on Quora whether there is a way to make a futures market for airplane tickets. It’s an interesting idea, and in principle I see no reason why it couldn’t be done. So I did a bit of Googling, and it turns out that a company called Skytra announced such an initiative. Nasdaq’s announcement is here. It reads, in part:

Skytra was recently established to help the $1 trillion per annum air travel industry hedge its revenue risk through the trading of cash-settled futures and options contracts based on a series of Skytra Price Indices. This is the first time that the industry will have financial risk management instruments available to them specifically to address revenue volatility.

You can read a bit more from my Quora answer.

Let’s unpack this a bit. Airlines’ revenue is famously volatile, irrespective of pandemics. When pandemics are not raging, revenues are affected by economic cycles, terrorism, tariffs, security requirements, alternatives to flying, etc. If airlines can sell futures in exchange for cash today, they can smooth out their revenues. Counterparties, to whom they sell the contracts, can lock in preferential rates for future travel. In theory, everyone benefits.

But this got me thinking: creating a futures market is essentially a data analysis play. You’re taking a company’s top line in its income statement—revenue—and hedging the risk that quarter-over-quarter volatility creates. This is data analysis: run an algorithm over a set of data, price the instruments correctly, and sell them through a partnership with Nasdaq.

There are a lot of other lines in a company’s income statement, of course. And here’s the kicker: for large companies, each line in an income statement is aggregated on the basis of thousands, sometimes millions, of transactions that occur every quarter. In other words, there’s a lot of data lying around inside companies’ accounting systems, not doing much of anything. This sounds like a classic data science opportunity!

In pre-data science times, private equity firms would often go through a company’s income statement line by line, and optimize the company’s performance by reducing expenses, recapitalizing the company (this is what a leveraged buyout is), and otherwise making the company more profitable for shareholders. But that’s also a slow, subjective, and expensive process.

I imagine—and here’s where the handwaving starts—that there must be some way for a company to collect all the data on the transactions that it incurs each quarter for given line of the income statement, and run an algorithm over the data. Which algorithm? I’ve no idea. With what purpose? I’ve no idea—other than to optimize the company’s performance. I warned you that I’m handwaving here.

So: who’s doing this? Are any SaaS companies selling this as a product? Are any companies doing this with their accounting data? Is this a terrible idea? If you have any knowledge about this, I’d love to hear about it.