Hacker News Reader

Comment by vermontdevil | Hacker News Reader

This is ridiculous:

Intellectual Ventures has always maintained that “traditional accounting rules don’t accurately reflect the value of our patent portfolio.”

Hate when companies not doing well try to dodge by blasting "traditional accounting rules" when these rules have been crafted over the years to accurately reflect the value of said company.

Bottom line - the returns for both funds are in the negative. Can't avoid this by hoping for some 'modernist' version of accountancy that might make the fund look good.

View on HN

"Hate when companies not doing well try to dodge by blasting 'traditional accounting rules' when these rules have been crafted over the years to accurately reflect the value of said company."

I agree that criticism of traditional accounting methods often points to inherent risks and even corruption, see Enron.

However, as I understand it, measuring firm value is not the primary goal of accounting. The way depreciation is calculated, for instance, may be convenient for other reasons, but it doesn't say much about real values at all. Accounting numbers are quite artificial.

When analysts calculate firm values, through methods like DCF, they use accounting numbers from the financial statements as their inputs to calculate cash flow etc., but that's just for lack of a better alternative. In fact the numbers don't make sense by themselves and many of the accounting rules just make the estimation of firm values more difficult.

> Accounting numbers are quite artificial.

At least three of these accounting numbers are very real: revenue, expenses and assets. When the dust from handwaving settles, they ultimately describe the health of a company any given fiscal quarter. Observing their first derivative values would typically show the direction the business is heading. It's this information that the crude corporate PID controllers (aka board meetings) use to steer the enterprise.

Look at the immense scepticism Tesla is met with based on operational stats, despite having actual, in-demand market product. Talk is cheap and accounting has been a great performance predictor many times over.

Almost all of Intellectual Ventures assets are patents with unclear value. Maybe they can use them to extract millions of dollars of settlements. Or maybe they're worthless.

When that's the majority of your balance sheet, whatever method you chose will have legitimate criticism.

And maybe my $20 pickaxe can be used to mine gold nuggets?

They can be valued with at least their acquisition value, which is a start. For all we know the "true" value (come liquidation) may now be zero, but still it puts a practical cap to how much IV can hype its business model.

And yet it is kind of true. Consider that patents in their portfolio are exactly equivalent to unexpended ordinance, ready to go when the opportunity presents itself. How do you value that? A patent my go its whole life and never be infringed, or it might turn into a $500M payday from an Apple or a Microsoft.

I don't think it's quite right to think of them as unexpended ordnance. They're more like signposted mines. In some ways they're only useful if they cover all possible paths through a solution space, since people can just avoid the minefield. How much overhead is there in trying to cover all the paths?

Even if you're just trying to catch out the unwary (or explicitly ignorant) who don't look for the mines, coverage of most routes is still pretty important.

This joint nature may affect value non-linearly, where A or B on their own are only worth X, but A and B together are worth 100X, or more.

This is not true, because almost no one reads patents. It's cheaper to do that almost all of the time than to always read and avoid them. Especially when you're avoiding patents that get invalidated in the end.

The only way to accurately value a patent portfolio would be to sell minority stakes in it on the open market. But if you expect to hold the patents until expiration and aren't short of cash then it's hard to see how valuation even matters.

Why assume the market is correct? Most people have no idea about the valuation of securities and patents are even harder to value

It doesn't matter what most people think. Only sophisticated, qualified investors are even able to bid on assets like fractional patent ownership.

As for whether the market is "correct" that's a meaningless question since the answer can only ever be determined in hindsight. But so far no one has been able to find a more correct approach.

The only thing that matters is some LPs having to mark their holdings to market!

But they have 95k patents and have done or backed scores of lawsuits with their $3b of capital. Surely that's enough! Y Combinator didn't need $3b to show good results despite regular VC having similar long-tail or hit-driven results.

All true, but how do you put an asset value on a particular patent? So if you've got a $100M fund and it spends $50M on acquiring patents and then the other $50M on prosecuting them as it can, it isn't possible to predict if one of the patents will be a gold mine or a dud. I think the closest thing I can imagine that would be like this would be mineral rights contracts. You spend $50M on mineral rights contracts on some amount of land and the rest on trying to develop useful resources out of those plots. If one has some easy to reach dense and salable minerals it pays for the others. But a plot that hasn't been geologically surveyed? What is it worth?

Well, the value would seem to be somewhere below $3b, so that's a good place to start for assigning the average value over those 95k patents...

If your argument is that there's no way to predict the future value of a patent, does it really have any value at all? If the patent game is basically a slot machine, investing in it shouldn't carry any expectation of return.

Perhaps, it would be like lottery tickets as assets. But the point was that it could certainly be something that isn't covered well, or at all, by standard accounting methods and practices.

the price of gold could go up a bunch at any moment too. this isn't unique.

Absolutely.

I think accounting rules often cause divergences from reality, but this is 10 years in and this fund has had AWFUL performance.

It's not totally ridiculous though. Patent valuation is a bit of a black art at the best of times. And on top of that, a new opinion from the Supreme Court or the Federal Circuit could instantly cause wild swings in value. There's no simple way to account for that.

Buffet complained about them recently too, do you think Berkshire's latest report accurately reflects their value?

Reply on news.ycombinator.com