How to short Uber (and other private companies)

I’ve been wondering how to short Snapchat since they passed up on Facebook’s offer to acquire them and raised capital privately which valued them at $3 billion. My curiosity was driven in part because AOL Instant Messenger, MySpace, and MSN Messenger were all once worth tens of billions of dollars before ultimately going out of fashion, and I believed that Snap wasn’t immune to this fate. I also naively have never understood the value prop of their product even though an entire generation of millennials seem to love it. Thankfully I never actually bet against them at a $3b valuation. And with Snap now a public company, the price will be set more accurately by market forces. But with other privately held companies like Uber being valued at a $68b, Theranos at $9b, and Zenefits at $5b, my curiosity continued.

One summer morning in 2015, I was having breakfast with a friend who works as an institutional sales-trader sitting on an equity derivatives desk at a major investment bank. He was telling me about how many of their tech clients would call them to try and structure hedges for their private company portfolios in order to protect themselves from a potential downswing in the tech market. Unfortunately, this was such a bespoke contract for the bank to structure that they would purposely over-charge their clients to enter them into such trades. I asked why there wasn’t a standardized way to do this, and referenced the credit default swap market as an example.

He said it’d be impossible to do so.

Since then there have been several detailed write-ups on this topic here, here, and here.

This is the story about how we made it possible, while also outlining the challenges and path we took to do so.

Venture Default Swaps

There are 5 elements to creating this product while complying with SEC regulations

1. Regulatory and legal restrictions
2. Market structure: counterparties and market participants
3. Collateral Management and circumventing stock transfer restrictions
4. Contract structure
5. Marking to market

To start off, one of our investors introduced us to Arthur Levitt, a former Chairman of the SEC, and we brought him onto our Board of Advisors to help us navigate the regulatory path. We then had multiple conversations with the SEC as well as with regulatory counsel on how this would be regulated. The SEC classifies these transactions as security-based-swaps, which are derivative contracts that are only available to eligible contract participants, or institutions that have at least $10 million in liquid assets on their balance sheets. Right off the bat, that informed us that our audience would be limited as we weren’t allowed to go to retail investors or even accredited investors since this would be a product that was only legally consumable by VC’s, hedge funds, family offices, and ultra high net worth individuals.

Over the course of the following months, we met with 65 of these prospective counterparties and narrowed down the major use cases of our contract to 3 core propositions for counterparties:

Group A: Hedgers

Counterparties that owned the underlying stock (most likely from an earlier investment stage in the company where they had been marked up substantially) that were interested in hedging their investments, effectively locking in a floor and a ceiling on their position.

Hedgers typically want to be able to post stock as collateral (which is challenging to do given pledging restrictions set by stockholder agreements) while also wanting to receive the proceeds of the trade immediately.

Group B: Net-longs

Counterparties that were interested in getting net-long in the underlying company, either to add to an existing position that they already had, or to get access synthetically due to not getting access in the primary rounds of financings.

Net-longs typically want an unlimited upside and long term expiries.

Group C: Naked-shorts

Counterparties that were interested in getting net-short in the underlying company, mainly due to a belief that the company was a fraud or just mis-priced and overvalued.

Naked-shorts typically want short term expiries and capped upsides for the longs; e.g. they don’t want to be on the hook for a stock that all of the sudden gets marked up 5x in the private market. They were also at the biggest information disadvantage considering the companies they were interested in shorting were privately held and not required to publish performance metrics.

Legally, all counterparties involved in a transaction would either (1) agree to not be acting on any material non-public information, or (2) acknowledge that they’re entering into an agreement where it’s likely that the other side of the transaction does in fact have material non-public information, and that they’re ok with it.

We then spent nearly 1 year doing customer development to come up with a contract structure that would fit in with all 3 sets of counterparty requirements and comply with regulations. This was challenging to say the least, especially given that the 3 sets of counterparties had vastly different structure requests in mind.

Of the 65 counterparties we met with, 56 said they’d be interested in the market pricing data, 35 said they weren’t initially interested in the top companies (Uber, Airbnb, and Snap) but would be interested in subsequent auctions for smaller Unicorns (which would have less two-sided interest), and 13 took an active interest in the top 3 names that we were running auctions on.

What we landed on was a structure with a long-term 10-year expiry that effectively enabled the long side and short side to enter into a contract with event-driven triggers, where if there was an IPO, M&A, or bankruptcy event in the underlying company, we would use the price of common stock at that liquidity event to settle the contract. To reduce counterparty risk, we required full cash collateral upfront to be escrowed with a third-party bank for the duration of the contract. Though this made it challenging for naked-shorts as they didn’t want to indefinitely lock up collateral, we alleviated the pain for the most natural market participants — the hedgers — by partnering with the lending teams at several major investment banks who would offer trade financing to the hedging counterparties, in addition to a personal loan based on how much underlying stock they owned.

From this structure, we created an ISDA agreement and circulated it amongst the dozen counterparties who expressed the most interest, comprised mostly of long-short equity tech focused hedge funds. Once we received their final commentary, we standardized it and finalized it. Since most hedge funds need to mark their positions to market on a regular basis for reporting purposes, we partnered with a valuation firm to help us mark these illiquid swap positions on their books.

(We considered other structures as well, such as a credit default swap or an insurance-type structure, but concluded that we wouldn’t have sufficient two-sided interest in such structures as the long side would be limited to needing to have a very large balance sheet; the only counterparties who could write these swaps would have to be major investment banks, who didn’t want to get involved given their interest in winning the IPOs and taking the companies public, and reinsurance companies, for which this was mostly out of their wheelhouse.)

Most of the interest we received was one-sided interest on the top dozen Unicorn names (e.g. everyone wanted to short Theranos, Dropbox and WeWork, or go long Airbnb and Snap).

The most controversial name that had sufficient two-sided interest in it has been Uber, of which we’ve collected over $200 million in interest. Over the last several weeks with the media’s negative headlines on Uber, we’ve gotten much more short interest while the long side has been getting cold feet.

Though we’re now getting to the point where our first transactions can get done, this process has made us question the viability and scalability of this market.

So if it’s possible, why hasn’t it been done?

Given the complexity of this project, there have been many forces to balance, and it’s a tough mix to fine tune. The reality is that this has been a challenging project on so many levels. It’s not like building an app — there is the regulatory component, the legal component, the product component, the clearing component, etc., and each one of these creates its own nuanced challenges. We think of them as circles in a Venn diagram, where each added component ends up as an additional a circle… pretty soon, the intersection of the circles at the middle — the market of people who will trade the product — gets very small because of all of these constraints.

While dealing with extremely sophisticated clients that have very high expectations, where there are only a handful, and where they are hard to reach, hard to satisfy, and have short attention spans, I would often ask myself what we could do to change the trajectory of the project to ease some of these frictions. The reason we worked on this was that it fundamentally stroked something deep inside us, from an intellectual love of doing something that couldn’t be done to creating market efficiencies in the private market as a result. The reality is that the notion of shorting private companies was a long shot. And though we have in fact made it possible to do so, there are so many structural obstacles to overcome, that long-term success — measured by market scalability — becomes highly unlikely.