[mikeyouse]

This isn't quite right -- The 4% loss could just represent the delta between what was already priced in and the actual loss. This case has been a long time coming so the expectation of the penalty should have been priced in long ago.

Secondly, and it depends on the jurisdiction, but penalties and fees associated with wrongdoing generally aren't tax deductible.

The US clarified this during the TCJA but most countries agree: https://www.clearyenforcementwatch.com/2018/02/1962/

[repsilat]

There's also that weirdness of financial companies having price-to-book less than one... (For UBS it's around 0.9. Don't ask me how it makes sense.)

[cm2187]

There may be good reasons for that.

First, there are two ways to value companies, depending on what kind of companies they are. The two extremes are on one side a consulting firm, where the value is worth only the present value of the future fees to be earned by the consultants. But the company itself has no asset, very little debt, and its balance sheet is pretty much irrelevant to its valuation. The other extreme is an investment fund, where the company is worth exactly the current market value of its assets, it has no other future revenues other than the expected incomes from these assets.

A bank is somewhere in the middle. It is kind of an investment fund in the sense that all its assets are financial assets that can be sold pretty much at their book value (unlike a factory where the book value may mean very little in term of how much the asset is really worth). But the book value may differ from the fair market value (that's often the case if there are bad loans). But part of the value will also be generated from profits that are unrelated to these assets: fees on payment processing, advisory fees, trading income, etc. So part of the value is also a PV of future revenues.

And these future revenues may be negative. If you look at the past 10 years, banks have experienced huge revenue volatility. From trading losses, fraud (Kerviel style), fines (like UBS here), bad loans (RBS and HBOS style), etc.

And then you can have all sort of weird technicalities. Like the book value may be increased for gains on own credit (if the company fair values part of its liabilities). That's a paper gain that the market is unlikely to give any credit for.

[calpaterson]

Presumably that is down to straightforward doubt as to the value of UBS's book. It is not uncommon for unpopular companies that are holding a lot of their balance sheet at "market value" (as banks do) to trade at discounts to their stated book. Take a look at LSE:LAND, a "£10bn" REIT trading at a £6bn market cap.

[mruts]

Well, just as this case demonstrates, certain companies have exposure to other things (like paying a gargantuan fine) that would make their shareholder equity trade at a discount.