Thread by @patio11

I don't know whether it's the contrarian in me or the part that is just fascinated by the mechanics of finance, but I feel like I would weight more highly commentary about the Vision Fund if it were followed with "And that's why I shorted it, using publicly traded instruments."

"How would you do that?"

The SoftBank Group is a holding company. Its ticker is $SFTBY. As a holding company, it is essentially just a math formula plus leverage. They publish the formula and the leverage.

c.f. the snapshot on page 31 of their 2019 AR cdn.group.softbank/en/corp/set/da…

Alibaba ($BABA among other places), SoftBank (9984), Sprint ($S), and Yahoo Japan (4689) are relatively easy to acquire on the open market. ARM is not. The Vision Fund is not. There's some minor residual value to the SoftBank Group.

To short the equity of the Vision Fund you short SoftBank Group, go long on those four constituents, and then either a) call it a day or b) figure out some way to go synthetically long on ARM.

This is *absolutely not* a recommendation to do this; this is a napkin sketch of how it could be done, if one believed one was much better informed than the market about the probability distribution of future returns to the Vision Fund.

"What does 'synthetically long' mean?"

We're synthetically shorting the Vision Fund here; creating a position which approximates the returns of a position which is easy to describe ("Short the Vision Fund") but not conveniently describable in a market instrument.

How would one go synthetically long ARM? I don't know; ask an actual professional in this. My guess, to a first approximation, is go long a semiconductor index and accept that there may be some outperformance of ARM versus semiconductors generally.