# What on earth is Barry Ritholtz going on about…

December 10, 2012 20 Comments

…in this post?

When we don’t know what any future outcome will be, but we understand the probability distribution — think of dice or a multiple choice exam — we have risk, but we do NOT have uncertainty. We never know what the roll of the dice will be, but we do know its one of six choices.

Huh? Look.

1. Risk IS uncertainty. That is the definition of risk. Please let me know if there are any questions.

2. If you know the distribution, you know the distribution, but that doesn’t make it non-uncertain. The distribution is perfectly known in a die roll, or in craps, but nevertheless if you make a bet on those things, your money is *at risk*, because you are not certain what the outcome will be. That is how normal people would talk anyway.

3. Perhaps more to the point: in financial markets, Ritholtz’s alleged forte, one *never* ‘knows the distribution’. There are times when it may be convenient to act as if you know the distribution, or some approximate distribution – you hedge a derivative with hedge ratio X because you *think* the prices of the derivative and its underlying would move together in that ratio, not because you ‘know’ it. This means that, by Ritholtz’s construction, the markets *have no ‘risk’* – just ‘uncertainty’.

None of this makes sense. So what is he really on about? Who is he really arguing against in his head, and what assertion(s) did they make, that annoyed him enough that he felt the impulse to try to define risk in such a weird and idiosyncratic way like this? In other words, what makes him think that trying to force there to be a distinction between ‘risk’ and ‘uncertainty’ was an important rhetorical tactic?

Speculations welcome. Actual author explaining himself, more than welcome.

UPDATE: I think I’ve gotten confused in comments trying to untangle the semantics (since I basically use the words interchangeably). Basically Ritholtz isn’t denying that a die-roll is risky, he’s saying we ‘don’t have uncertainty’ if we know the probability distribution – just risk. Two things. 1. While it’s true we ‘don’t have uncertainty’ *about a die-roll’s probability distribution*, that’s not the same thing as having no uncertainty about its outcome. The fact that we speak in terms of probabilities is precisely because we are uncertain of its outcome. Again I ask the question, would Ritholtz feel the need to hedge such an outcome? You don’t have to hedge certain events. (Actually, you can’t; no one would sell you such a ‘hedge’.) 2. ‘Knowing the probability distribution’ is not something that ever happens in the markets. Even on the die roll, how do you know the die is fair? You assume it? Hmm, people assumed house prices couldn’t go down. Since in reality, the world is always ‘uncertain’, risk and uncertainty are identical in practice, even by Ritholtz’s standards.

You’re singling out this guy but a distinction between risk (quantifiable) and uncertainty (unquantifiable) is widespread. Or is it the idea that risk is quantifiable and we know the quantity, vs uncertainty which includes both unquantifiable probabilities and ones which are quantifiable but we don’t know what the quantity is?

Is this a reasonable or useful distinction? I don’t know, but it doesn’t sound prima facie retarded. But then most persistent bad ideas don’t.

The distinction between aleatory uncertainty (the outcomes are random) and epistemic uncertainty (the outcomes and/or their distributions are unknown) is useful. But calling the former “not uncertainty” is non-standard vocabulary, and calling the latter “unquantifiable” is just incorrect.

Sure, the quantifiable/unquantifiable distinction is reasonable. But I would say both *are risk*. There is quantifiable risk and unquantifiable risk. You don’t call the latter ‘not risk’, that makes no sense.

Except Keynesian woo. Sounds retarded; is retarded.

well you probably noticed that Barry started with his interpretation of Mouboussin’s definitions:

“In other words, his view is that the future is always unknown — but that does not make it “uncertain.” ”

which I think is nonsense. If it’s unknown, by definition, it’s uncertain. Which is, of course, not to say that the distribution of possible outcomes is undefined.

Kid Dynamite —

Consider what you are saying: Isn’t the future by definition always “unknown”? By your statement, doesn’t that mean there is always Uncertainty?

That make the use and overuse of the word (as you define it) utterly meaningless. If there is alwayss uncertainty, how can you blame so much on it this month/quarter/year ?

MM defines it in a way that is useful.

“doesn’t that mean there is always Uncertainty? ”

absolutely. and the level of uncertainty changes – frequently.

Barry – if we roll a die, what will the outcome be? this was one of your examples. You claim that because we know it’s 1 of 6 outcomes, that we don’t have uncertainty.

I counter that claim by saying that you should be willing to bet your net worth any time you are in a situation with no uncertainty. Do you want to bet your net worth that you can tell me which side the die will land on on the next roll? Of course not – because there is uncertainty.

Yes, I will bet you that the out come will be 1,2,3,4,5,or 6

That much is certain.

As to any given roll, well, we know the statistics — what odds are you willing to give me to get me to assume the risk of an unknown (but not uncertain) outcome ?

It is certain that the die will roll something between 1…6 and it is certain that a stock price will be a real number greater than or equal to zero. But this doesn’t mean either thing ‘has no risk’. Do you really not understand this or are you intentionally being obtuse to make some rhetorical point?

It’s nice that you know the statistics of a die roll (unlike a stock price) but that doesn’t convert the event from risky to non-risky. The fact that the outcomes follow a nontrivial probability distribution (whether we know what it is or not) means it is risky. Knowing what the probability distribution is just means you know how to hedge it (which is why you’d ask for odds). You wouldn’t have to hedge it at all if it were ‘non-risky’ so again your statement here is an admission that your earlier framework makes no sense.

ps – Barry – I would probably agree with you that journalists who write “Stocks down on uncertainty over fiscal cliff resolution” are being silly, but hey – we both know that journalists frequently write silly stuff.

What they probably really mean is “the uncertainty around the outcome of the fiscal cliff negotiations results in potential price swings which are larger than the price swings accompanied by normal levels of uncertainty in the market. the current level of uncertainty is thus higher than the normal “average” level of uncertainty”

but that doesn’t have quite the same headline ring on Yahoo Finance.

Yes, I basically just should have said ditto to your posts KD.

Simple question for Barry: suppose he inherits a book with one trade in it, a binary option paying out $1 billion if a specific die to be rolled on 12/20/12 comes up #5. After examining this book, does he go and tell his boss, “Don’t worry, there’s no risk in this book”

??

Barry:

Okay, so there is always some uncertainty. Some sources of uncertainty are more quantifiable/predictable than others. I guess my real question is, what bothers you so much about saying this?

simple proof of my side of this semantic argument (And I do think it’s a silly semantic argument):

what will AAPL’s closing price be today?

my claim: AAPL’s closing price today is uncertain.

even though we know with certainty the set of possibly outcomes (0, .01, .02, .03 … 500, 500.01, 500.02, 500.03…. 867, 867.01, 867.02 etc etc etc)

$529.82

But we always knew it was going to be a dollar amount, between $0 and a $ million.

Using Mauboussin’s definition in terms of probability distributions, we know it was “unknown, but not uncertain.”

Exactly.

He’s talking about the classic distinction between risk and uncertainty made by Frank Knight. This distinction may not be a good one, but Knight is the one who came up with it decades ago.

Read these two — and see if it doesn’t clarify that there is a distinction between the two.

Previously:

Kiss Your Assets Goodbye When Certainty Reigns (Bloomberg, November 9, 2010)

http://www.bloomberg.com/news/2010-11-10/kiss-your-assets-goodbye-if-certainty-reigns-commentary-by-barry-ritholtz.html

There’s nothing new about uncertainty (Washington Post, July 7 2012)

http://www.washingtonpost.com/theres-nothing-new-about-uncertainty/2012/07/06/gJQA5hnmUW_story_1.html

So does this really boil down to the fact that the way ‘uncertainty’ has been used in layperson financial reporting has been getting on your nerves? If so:

1. I’m sure there has been some lazy/stupid usage of that term,

2. I don’t get why it would bother you so much *shrug*, but either way

3. That’s not a good reason to try to redefine the words in ways that make no sense.

I asked this above:

If you inherited a book with only a trade that was a binary option on die roll=5, would you go around saying that book had ‘no risk’ in it? and make no attempt to hedge it?

Let me know,

you read this past post of mine, right?

http://kiddynamitesworld.com/when-you-cant-quantify-your-risk/

I use “risk” and “uncertainty” in largely similar fashions, although not identical.

Pingback: The Jihad Against ‘Uncertainty’ Continues « Rhymes With Cars & Girls