Take Your Lumps

This post is by Fred Wilson from AVC

Click here to view on the original site: Original Post

The Gotham Gal and I went through our (actually her) angel investments yesterday and figured out which ones went under in 2018 so we could take the tax write-offs on our 2018 returns.

It is an odd exercise. Kind of like reading the obituaries.

But it is an important exercise for several reasons.

First, taking the write-offs against the gains shelters the gains so they can be re-invested in full. Over her first six years of investing (2007-2012), she has realized a bit more than she invested and the losses have sheltered the gains so all of that capital can be reinvested. And the investments that remain unrealized from that cohort are all solid now and will likely produce another 2-3x on invested capital.

But it also a nice “post mortem” process to go through the ones that didn’t work and think a bit about what went wrong. We don’t

about the losses, but taking some time to run through them is helpful.

Sometimes failed investments turn into the “living dead” in which you end up a tiny investor in another company by virtue of an acqui-hire, a distressed sale, or some other such transaction. It is generally a smart idea to sell your stock back to the company or another shareholder or abandon your interest and take the loss on those kinds of investments. The tax loss is often worth more than the stock you own. A regular process of going through the losses will surface opportunities like that too.

The bottom line is that angel investing is risky business. Super early stage investing, like the kind the Gotham Gal does (she is most often the first check into the company), will produce loss ratios of 50% or higher. The winners eventually bail you out and super early stage investing ought to produce 3x on capital or better (or you shouldn’t be doing it). One nice advantage of this model is the losses come early and the wins come much later. Taking your losses, getting the write-offs, and sheltering your gains is an important part of the model and it is best to have a regular process to make sure you are taking the losses when you can.