Sunday, March 30, 2008

Hypothesis testing

Andrew Gelman points to an article on The Fallacy of Hypothesis Testing. I'm not sure I agree with everything in the article but this paragraph caught my eye:
Third, I've learned that the scientific community's emphasis on hypothesis-based research leads too many scientists to devise experiments to prove, rather than test, their hypotheses. Many journal submissions lack any discussion of alternative competing hypotheses: Researchers don't seem to realize that collecting data that are consistent with their original hypothesis doesn't mean that it is unconditionally true. Alternatively, they buy into the fallacy that absence of evidence for something is always evidence of its absence.
Gelman responds:
... I imagine many of my social science colleagues could present a defense of hypothesis testing. (Just to be clear, I think we're talking here about the idea of posing and testing hypotheses, not the textbook statistical methods called "hypothesis testing." The hyp testing that Pepperberg is talking about could just as easily be done using confidence intervals or whatever; her real distinction, I think, is between studies that are exploratory and studies that are designed to test particular scientific theories.

The general criticism seem to be that hypothesis testing is conducted in the absence of competing models. But if different models lead to the same hypothesis test then the question seems to be one of differentiating between alternative models.

Harpers Ferry

We were at Harper's Ferry most of this past week. Quiet, peaceful ... it looks to be a wonderful view in the fall.

Monday, March 24, 2008

Administrator/Teacher ratio

MR points to an InsideHigherEd article:
...the federal statistics provide evidence for another shift, in which the majority of full-time professional employees in higher education are in administrative rather than faculty jobs.

This sounds like K1 and K2's school. Looking at the year book it seems almost like there is a 1-1 ratio of teachers to administrators. I wonder where the fees are going.

Langkawi

Ondine Cohane's article on Pulau Langkawi makes it sound like an idyllic place and it certainly could be given that she spent her time at The Andamman, The Datai and The Four Seasons. I had looked into some more affordable places (like Hilton/Sheraton etc.) but the reviews at Tripadvisor were so mixed that I decided to back off from it.

Some good links to the credit crisis

Some good links to the credit crisis:

1. How The Fed took the money out of monetary policy:

Here’s how a repurchase agreement would change the Fed’s balance sheet, after offsetting it with an open market operation:

Changes in the Fed's balance sheet after a $1,000M repurchase agreement, offset by an open market operation
Assets
US government securities
-1,000
Repurchase agreements
+1,000
Reverse repurchase agreements
0
Direct loans
0
Other assets
0
Liabilities
Currency in circulation
0 (-1,000 + 1,000)
Reserve balances
0
Source: Federal Reserve, H.4.1 release.

By themselves, TAF loans would increase both assets and liabilities of the Fed, just like open market operations and repos. But, once again, the Fed partially offset those loans by selling securities and withdrawing cash from the system. Here’s the simplified balance sheet on December 26 and August 15:

Federal Reserve's balance sheet, $ millions
Assets
Aug. 15, 2007
Dec. 26, 2007
US government securities
789,601
754,612
Repurchase agreements24,00042,500
Reverse repurchase agreements-31,941-40,542
Term Auction Facility loans
0
20,000
Direct loans2644,535
Other assets37,05852,869
LiabilitiesCurrency in circulation813,085829,193
Reserve balances5,8974,781

Source: Federal Reserve, H.4.1 release.

The first one is the Term Securities Lending Facility (TSLF), to open on March 27. At this new window, all primary dealers -all banks and brokers that trade in government securities with the Fed- are allowed to borrow up to $200bn of government securities for 28 days. Borrowers must pledge collateral for these loans, but the minimum quality of the assets is even lower than for the TAF (it includes federal agency debt, federal agency residential-mortgage-backed securities (MBS), and non-agency AAA/Aaa-rated private-label residential MBS).

The second institution is the Primary Dealer Credit Facility (PDCF), which started operating on March 17. This venue provides overnight loans to all primary dealers, backed by even riskier collateral: they accept all collateral eligible for repos, plus investment-grade corporate securities, municipal securities, MBS and asset-backed securities. With the PDCF, all primary dealers have de facto access to the discount window, from which only depository institutions could borrow before.

Here’s the balance Fed again, in December and after the PDCF opened:

Federal Reserve's balance sheet, $ millions
Assets
Dec. 26, 2007
Mar. 19, 2008
US government securities
754,612
660,484
Repurchase agreements42,500
62,000
Reverse repurchase agreements-40,542-46,143
Term Auction Facility loans
20,000
80,000
Primary Dealers Credit Facility
0
28,800
Direct loans4,535
125
Other assets52,869
36,603
LiabilitiesCurrency in circulation829,193818,362
Reserve balances4,781
3,507

Source: Federal Reserve, H.4.1 release.

With its new tools, the Fed has provided liquidity without printing much money. It has temporarily absorbed risky and illiquid securities, and supplied government securities, which are risk-free.

2. Mark Thoma's response to Steve Waldman's Credit Crisis for Kindergartener's:

"Couldn't the parents force the kids to keep their promises (under threat of a large penalty for default)? Either do what you promised, or incur some punishment that makes doing the chores the only reasonable choice? That seems a lot like the way a court would enforce contracts, so we need one of the kids to declare bankruptcy (or simply refuse to work and accept the punishment of having assets stripped, getting sent to their room, grounded, etc.) to get this going."

JP Morgan's decision to raise its price for Bear Stearns from $2 to $10 weakens the penalty effect.


Wednesday, March 19, 2008

Voter theories

Louis Menand's New Yorker article The Unpolitical Animal was a fun read. Some interesting sections:
1. Skepticism about the competence of the masses to govern themselves is as old as mass self-government. Even so, when that competence began to be measured statistically, around the end of the Second World War, the numbers startled almost everyone. ... About forty-two per cent of voters, according to Converse’s interpretation of surveys of the 1956 electorate, vote on the basis not of ideology but of perceived self-interest. The rest form political preferences either from their sense of whether times are good or bad (about twenty-five per cent) or from factors that have no discernible “issue content” whatever. Converse put twenty-two per cent of the electorate in this last category. In other words, about twice as many people have no political views as have a coherent political belief system.

2. In a paper written in 2004, the Princeton political scientists Christopher Achen and Larry Bartels estimate that “2.8 million people voted against Al Gore in 2000 because their states were too dry or too wet” as a consequence of that year’s weather patterns. Achen and Bartels think that these voters cost Gore seven states, any one of which would have given him the election.

3. The most widely known fact about George H. W. Bush in the 1992 election was that he hated broccoli. Eighty-six per cent of likely voters in that election knew that the Bushes’ dog’s name was Millie; only fifteen per cent knew that Bush and Clinton both favored the death penalty. It’s not that people know nothing. It’s just that politics is not what they know.

4. There is nothing in the Constitution requiring candidates to be listed on the ballot with their party affiliations, and, if you think about it, the custom of doing so is vaguely undemocratic. It makes elections a monopoly of the major parties, by giving their candidates an enormous advantage—the advantage of an endorsement right there on the ballot—over everyone else who runs. It is easy to imagine a constitutional challenge to the practice of identifying candidates by party, but it is also easy to imagine how wild the effects would be if voters were confronted by a simple list of names with no identifying tags. Every election would be like an election for student-body president: pure name recognition.

Monday, March 17, 2008

Update on Can the Fed engineer a turnaround

Update on this post where I suggested:
Assuming that the same institutions who invested in these securities can classify them as toxic or possibly non-radioactive then a workout perhaps similar to the Resolution Trust for S&L can be created for the toxic securities ...

Today, Paul Krugman echoes something similar in the NYT:
Looking ahead, we probably need something similar to the Resolution Trust Corporation, which took over bankrupt savings and loan institutions and sold off their assets to reimburse taxpayers.

Today the news is all about the Fed "bailout" of Bear Stearns - the Fed has agreed to help finance the buyout of of Bear Stearns at $2 per share by JP Morgan by holding some BS MBS as collateral. Some details at MR and EV.

Update (3/19): JDH expains why this is not a bailout. I would tend to agree although:
"True, the Fed did offer a $30 billion non-recourse loan to JPMorgan to sweeten the deal. But what twist of logic would lead us to describe that as a "bailout" of Bear as opposed to an inducement to JPMorgan to help clean up the mess?" My response: How non-recourse is the non-recourse loan?

Why did the Fed force a fire sale of BS at $2 a share? Is it perhaps sending a signal to other banks/hedge funds/SIVs/SPVs to clean up their act or the Fed will do it for them on the Fed's terms? MR writes:
"The Fed's regulatory powers make crisis deals less than fair. If you, as a bank, don't accept the Fed's terms, you can be prosecuted or thrown in jail or at least ruined by your friendly regulator. Being an advocate of the rule of law, I'm not entirely comfortable with this arrangement, but it does mean that the Fed has a much easier time managing crises. Keep in mind also that the failing banks are indeed the most likely ones to have been criminal, so the unfairness is not usually being applied to the innocent."
MR continues:
"First, the very active role of the Fed in the Bear Stearns crisis must, in the long run, give rise to a fundamental revaluation of the role and powers of the SEC, the entity technically responsible for investment banks. The SEC now appears relatively toothless.
Second, the more commitments made by the Fed, the more we lose the (quasi) independence of our central bank; for a large commitment Treasury sign-off is needed. The realignment of the regulatory universe will eventually emerge as a big story from the current crisis, though it is hardly commanding much attention right now."