Irrationality is the real invisible hand | Hacker News Reader

Irrationality is the real invisible hand

technologyreview.com

2 pointsrazorburn4 comments

There was nothing irrational about the crash. For a long time, a lot of people where getting rich and having a good time spending. They were simply "rationally" exploiting the system as well as they could.

Somehow I still remember an article that the whole housing misery started with laws that required more credit to be given to poor people, resulting in schemes to make that bad credit viable (we all know how that went). Why is that not being mentioned anymore? It would indicate that instead of too few laws, too many laws started the problem.

That line of conservation disappeared b/c it was essentially attention-bait for people predisposed to believe it; upon any real inquiry (eg: reading the actual text of the CRA, understanding what it did and didn't do, and the total size of CRA-associated lending programs, not just reading an editorial or two) it falls apart as anything like a causal explanation; much like speculators in the financial markets opinion columnists -- speculators in both senses of the word -- rush to jump on a meme once it's climbing and rush to jump off once it's on the way down, often leaving a confused bagholders in their wake, wondering where all that strum und drang got to (like you, in this case)...didn't you get the memo?

If you want to point at past interventions that're more-directly responsible for triggering today's problems you might start by mapping out the intersection between:

(i) pension funds, the size of their assets under management, the legal restrictions on what they can invest in, and the performance window they need to hit

(ii) how fed interest rate policies can impact bond yields across the broader bond market

(iii) how securitization works, including some comparison (before securitization/after securitization) of attributes like yield and 'ratings', and in particular paying attention to drawing connections here with point (i)

That's a lot more work and it's still almost certainly not "the whole story", but there's a definite story there that (in fact) adds a lot of color to the picture, even if it's a lot less pithy than the CRA.

I see, thanks! I don't understand everything you are saying, but I still. I think rather than looking at individual laws and financial constructs, it would be more interesting to study why people accepted investments without transparency. I think more transparency might be the correct answer, rather than regulations as to what kinds of investments should be allowed for whom.

I remember reading about another stock market bubble (I think India Tea Company), where at one point someone sold shares for a "venture whose activity shall only be revealed later", who then disappeared overseas with the investors money. But really, if people are willing to invest in that kind of thing (0 transparency, extreme risk), can laws really help, and would they be the right answer?

I was actually being a bit of a smartass, but I've decided to repent. I don't have time to tell you the full story, but here's a fuller outline.

I assume you know what a bond is; if you don't, you'll need to look it up.

(1) The fed basically sets the interest rates in the bond market. The real mechanisms are complicated and have a lot of "moving parts", but the basic picture is:

- the fed basically sets the interest rate on new "treasury securities" (treasury bills, notes, etc.).

- "treasury securities" are considered 0-risk (b/c the borrower can just print the money to pay them, basically)

- when other parties want to raise money by selling debt (eg: offering bonds) they have to offer more interest than the going rate on "treasury securities" (b/c other parties will have higher than 0 risk of nonpayment, and consequently the higher interest is what attracts investors to buy the bonds)

- basically the relationship is like: the riskier the bond, the higher the interest rate (over and above the federal rate) it will offer

(2) throughout the mid-late 90s and into the early 2000s the fed held interest rate at abnormally low levels (at least by historical standards).

The net effect of holding interest rates that low for that long was that the range of interest rates you could get on bonds was lower than historical norms (by a few % points at least).

(3) There are enormous pools of highly-regulated money-for-investment held in vehicles like pension funds (domestically and also abroad). CF this chart (from 2006):

http://books.google.com/books?id=x4d-lCjRY8AC&pg=PA4&...

...to get some idea how much money is in USA-based pension funds alone (there's more overseas where pensions are higher).

The details very a lot between categories of funds, but generally highly-regulated funds who are more about risk-minimizing (like pension funds) than return-maximizing will try to invest heavily in bonds or bond-like instruments; again this is complicated but roughly speaking the reasoning is:

- bonds or bond-like instruments usually have a defined series of payments whose value is easy to calculate; the only real risk is that the borrower stop paying

- there's huge pressure not to default on bonds; a dropping stock price is bad for shareholders, but defaulting on bonds risks your being able to continue to operate (b/c you have no way of raising short-term funds)

...which makes bonds historically lower-risk, lower-reward compared to (eg) stocks. Pensions and pension-like funds have another reason for preferring bonds:

- pensions need to have cash on hand on a regular basis (to pay out claims)

- bonds always pay out on a regular basis (until they don't); to turn stocks into cash you have to sell them (or hope the dividends don't change on you)

Thus historically you'd have a huge diversified pile of bonds (mix of higher-return / higher risk and the opposite) as the major component of such a fund's portfolio (with various other investments tossed in).

The consequence of a long period of historically low interest rates meant that bonds no longer offered the same kind of returns; even a few % matters here, b/c if eg you are using an internal interest rate of 4% then going from a portfolio average of 7% to a portfolio average of 5% means you're actually going from 3% net of inflation to 1% net of inflation, etc.

Now, if these were just hedge funds or other speculative vehicles that'd be one thing; but, pensions legally are contractually obligated to make their payments, so failure-to-perform isn't just "too bad, so sad" but potentially legal nightmares; additionally, USA pensions are partially backed by an equivalent of the FDIC

http://en.wikipedia.org/wiki/Pension_Benefit_Guaranty_Corpor...

...which currently is self-funding but in a real disaster would probably graduate to explicit federal backing.

So summary: due to fed policy vis-a-vis interest rates over an extended period of time (well over a decade now) it's been harder and harder for pensions to meet their performance targets with their traditional strategies; due to their nature this isn't just bad news, it's possibly "against the law" in some sense.

(4) Enter Securitization

What securitization basically did was:

- take a 1000 mortages originating at about the same time (july 2004) but geographically distributed and with a range of credit scores / home values / interest rates

- set up an entity that receives all monthly mortgage payments

- as the money comes in, the entity first dumps it into "bucket 1"; when "bucket 1" is full you send that out to its owner(s), then you dump what's left into "bucket 2", send that out when it's full, etc, until the money runs out. It's possible that, eg, once a handful of people default or prepay their mortage that buckets 5 on down stop getting any money.

- you call the "buckets" tranches, and you do some "financial engineering" to decide how many buckets to put in and how big to make the buckets; depending on how you choose you typically wind up with some top-level tranches that have low risk but higher effective interest rates (compared to normal bonds with the same interest rates)

- the reason this works is that even though the bottom buckets are going to go to zero (almost inevitably) you're concentrating most of the risk in the bottom tranches and thereby the higher tranches have lower-risk

- zoila! you've created a bond-like instrument that has higher returns but the same apparent risk (ratings agencies had a huge role here, again too little time to detail in full)

So, there was basically a huge demand for "yield" (interest rates on bond-like products) but a huge shortage (due to fed interest rate policy); the whole securitization scheme came about partly to meet that demand.

The consequences of securitization are varied: it increases willingness to loan to high-risk borrowers (b/c you sell off the mortgage as part of the security, and having some high-interest rate, high-risk "pepper" in the security is essential to being able to offer high interest rates.

But for a lot of market actors the products looked really appealing as a way of "not going out of business".

There's a lot more you could do to draw out that story with enough motivation; it's far from the whole story -- or the only story -- but it's one of the underlying currents leading up to the current predicament.

It's nice to have a single thread to tug on that explains everything but that's hard to find; the financial world is a huge web of interconnected actors with interconnected motives.

Reply on news.ycombinator.com