1. Wikipedia Tourism #18

    March 30, 2009 by Craig

    From Cousin:

    In 2004, genealogists discovered that U.S. Presidential candidates George W. Bush and John Kerry shared a common ancestral couple in the 1500s. It was reported that the two men are sixteenth cousins, thrice removed. However, the two are in fact ninth cousins, twice removed. Also, in 2007, it was revealed that then-U.S. vice president Dick Cheney and then-Democratic presidential candidate Barack Obama are eighth cousins.

    That’s right folks: Obama (ie: the Black Guy) and Cheney are more closely related than Bush and Kerry (two Yalies).


  2. Wikipedia Tourism #17

    March 17, 2009 by Craig

    From Polar Bear

    [they] are nearly invisible under infrared photography.

    I would go see “Polar Bear vs. Predator”.


  3. Mark-to-Market for Dummies

    March 1, 2009 by Craig

    Marco talks about the economy and asks about a policy called “Mark-to-Market“:

    One of the laws that might get changed is mark-to-market accounting were you have to declare assets at the current market value. I am no expert on this–but I am told that if we allow you to declare some of these complex financial instruments at “what you think they are worth” we could get good mileage out of that because it would remove uncertainty. I am not an expert–and I don’t fully understand the value here. I’m sure someone can explain it.

    I’m no expert either, but I have a very good lay understanding of the issue, so I figured I’d write up my explanation.

    There’s two major factors used to value any particular debt at any given time. The first and simplest is it’s “present value“. This is the payoff amount adjusted for the interest paid. The math for this is well-understood. Example time: let’s say that you promise to pay $100K in 10 years. If I want a 5% return on my investment in you, I can run the numbers and find that this deal is worth about $60K as of today… that is, if I gave you $60K now and you gave me $100K in 10 years, I would have made 5% return on my investment.

    This doesn’t take into account risk though. If you fail to come through on your promise to pay, then I lose everything. I have to take this into account when determining the value of this loan to you. If I think there’s a 50% chance you’ll stiff me entirely, then I might reduce my “expected value” of this loan by 50%: I’ll only give you $30K instead of $60K for the same $100K/10 year promise.

    Determining risk is where the difficulty comes in. What’s your chance of default? I pulled 50% out of thin air; when it comes to real life the factors involved are numerous, complex, and often unknown. This is where financial magic (and potential for profit) comes in.

    Let’s say that Alice knows that you’re about to inherit $1M and can easily pay off your debt. She might figure your risk of default is only 10%, compared to my 50%. Accordingly, she might value the loan at $54K instead of my $30K. She can then offer to buy the loan from me for $42K. This looks like a good deal to both of us; I get $12K over my “expected value” of the loan and get some risk off my books. Alice gets a relatively reliable deal for $12K less than she thinks it’s actually worth.

    Or, you might have Bob who knows you have cancer and will need to spend your money on medical bills. He figures your chance of default is 90%. He values the loan at $6K. If I demand $30K to sell the loan, he doesn’t go near it with a 10 foot pole.

    Mark-to-market means that I, as the lender, have to value a loan at what everyone else will pay for it, regardless of what I think the real expected value/risk is. If Alice and Bob both make bids on my loan for $42K and $6K, I can record a $12K increase in the value of my asset even if I decide not to sell it. I could then use this as extra collateral if I wanted to take out a loan of my own.

    However, in 2008, Alice went away, leaving only Bob. Bob bids $6K for my loan asset. I still don’t want to sell it (especially at that price), but the law says I have to use his bid as the expected value. All of a sudden, my books show a $24K hole. (I’m in even bigger trouble if I got a loan based on your asset that I thought was worth $42K).

    None of this (directly) effects whether or not you’ll be paying me $100K in 10 years; that probability hasn’t changed. If my initial risk assumption was right, there’s still a 50/50 chance you’ll pay, and the real value of the loan is still $30K. But I have to treat it as if it’s worth $6K, simply because the only people wanting to buy it (Bob) thinks its worth that much.

    The proposed solution is to use something other than Bob’s bid to set the value of the asset; that’s where “what you think it’s worth” comes into play. The big problem with this is that we don’t have any measuring tool better than market prices to value risk. Risk assessment is so complex that usually we say “screw it” and let the wisdom-of-crowds do the work for us. Unfortunately, the crowds aren’t always wise / rational, and so mark-to-market might be hurting us now above and beyond the real problems in the economy.

    The other trouble with non-market-based approaches to value is that they’re easier to manipulate. If I think that the chance of default is really 50% but I say it’s 25%, I can artificially boost the value of my asset and use it as collateral. In this case, the person lending me money has made a riskier investment than he thought. Alternately, I could exaggerate the risk of default, claim a loss on the value of the asset, and use it to avoid some of my taxes. Mark-to-market is useful because, even though it was imperfect, it was at least transparent; you couldn’t get away with too much.

    An ideal solution would have both the transparency and objectivity of mark-to-market but not rely on the occasionally-irrational markets for its valuation. So far, I don’t know of any system that would provide this. However, necessity is the mother of invention; we might just see a better system emerge as a result of the crisis we’re in. We may even get stronger because of our illness.