2011 Nobel Laureates in Economic Science

This year, the award goes to

for their contribution to the research of macroeconomics!

Correlation vs. Causality

From Calculated Risk:

"Residential investment (RI) is the best leading indicator for the economy. This isn't perfect - nothing is - but RI is usually a strong leading indicator for the business cycle. The slump in RI helped me call the 2007 recession correctly, and the lack of a recovery in residential investment is a key reason the recovery has been sluggish and choppy so far."

What is the correlation stated? What is the causality stated? Explain how the causality is established? Does it base on a solid reasoning (theory?) or just come out of the middle of nowhere?

Assuming the statement is correct, what is the most likely policy recommendation? Is that going to work or is that going to hurt?

Philosophical Economics

Just a (few) piece(s) of cake.

Neither full understanding nor agreement necessary. 'Inspiration' ranks first.

Uhlig on economics and reality

Caballero on macro

Hamilton -> Bunnies on QE

I like what Uhlig says, " ... ... but in science, either put up or you shut up ... ... ". Unfortunately, this does not necessarily apply to other fields, e.g. politics (you bug up, don't shut up, and don't put up), relationship (either you or your partner shut up), arts (you put up and shut up) ... ...

Boom and Bust in Housing Market

今年頭 9個月，持貨一年內轉手的單位較去年同期激增 1.1倍，其中 84%單位屬 300萬元以下，顯示炒風已蔓延至中下價樓，「炒家利用亢奮氣氛，吸引市民入市」。

The quoted text reminds me of a paper of Burnside and et. al (2010).

http://www.econ.upenn.edu/system/files/moneymacro11102010.pdf

Burnside and et. al develop a model of boom and bust in house prices. To cut a long story short (as I always do), the model consists of the following features:-

- there are 3 types of agents, namely, infected, cured, and vulnerable, which correspond closely to "speculator", "intelligent investor", and "speculator-lured-to-be"

- in period t, t+1, t+2 ... ..., social interaction is allowed so that the infected, the cured, and the vulnerable, after interacting with each others randomly, may convert from status to status with some probabilities

- the infected has a belief that the property market will boom to state H (high state) with the timing of the boom not known

- agents with stronger beliefs are more likely to convert other agents to their beliefs

- agents take action (buy or rent property) after each round of interaction

- the market state (H or L) realized subsequently

Burnside and et. al. shows that the social dynamics would "produce a rise and fall in housing price" and they claim that the model can account for key features of the boom-bust cycle in US housing market (price booms are associated with the influx of new buyers).

Mulligan (2010) (see my previous post) proposes an alternative boom and bust story, but with reasoning differs from that of Burnside and et al. Nevertheless, there is at least one common feature between the two stories and that is "EXPECTATION MATTERS"!

By the way, why call the agents infected, cured, and vulnerable? Well, according to Rebelo, the paper's co-author, they are so named because the evolution of the fraction of population with different views is similar to that in Bernoulli's study of infectious desease spread in Europe.

I nearly laughed to death when I were told of the reason.

QE2

This is to those who want to know a little more about QE2.

James Hamilton gives a very detailed and comprehensive description of what he believes QE2 is at here, here, here, here, here, here, here, here, here, and here.

To cut the long story short, Hamilton addressed the following:-

1. why the FED, not the Treasury (in theory, the proposed objective(s) of QE2 can also be achieved through Treasury operation)

2. the mechanism of QE2 (if you are really interested in QE2, you can read Hamilton and Wu's paper here)

3. QE2 is not the one and only one policy tool the FED has

4. the main objective of QE2 is to avoid deflation (ibid)

5. in endorsing QE2, the FED should keep watching commodity prices changes closely

For ordinary people like you and me, 2 and 4 are areas where we "should" pay our limited and scarce efforts on as they have strong implication(s) on portfolio management. In case you still have time after struggling on 2 and 4, you may wish to learn a little more about 3 and/or 5. 1 is for nerds.

A digress: what I think is "missing" in Hamilton's series of posts are

1) a normative fiscal policy recommendation and/or

2) a discussion of the (short/long term) effect of government expenditure under QE2 on the overall economy.

Economic Analysis

I don't like the idea of dividing the study of economics into microeconomics and macroeconomics. For me, what matters are:-

1) what are the "all other things remaining the same" in models on which economists keep drudging; and

2) the reasons behind the assumptions

It is not a "crime" to hold quantity supplied constant or to have prices remain unchanged, but it is a "crime" to "assume" such for pure analytical purpose(s) / convenience.

Mankiw's Choice

From Statistical Modeling, Causal Inference, and Social Science.

Mankiw's original article.

Mankiw's "partial"/"incomplete" response to some of the "diatribes"

What's "wrong", if any, with Gelman's reasoning?

Is little knowledge dangerous?

To write (blog) or not to write (blog), that's your question. ("Obviously", I prefer to be paid to write (or paid not to write); after all I am not live-to-work type of person.)

Perhaps, Ed has a good "answer" to "blog or no blog".

Subsidized Housing

From today's Appledaily.

Trust me, the pricing of the monthly rentals and/or the pricing of the "subsidized" housing is a good PhD dissertation topic.

However, everything is at a cost, and this topic is very challenging. There are at least three issues to address, and they are:-

1. there is an option to purchase. You have to price this option and include it in the monthly rental or charge it separately in a lump sum (though I guess the gov't is going to give it free of charge);

2. the underlying asset is not an investment asset, so CAPM and forward risk neutral approach can hardly be used; and

3. the total rentals paid (during the renting period) are future purchase price deductible.


******

My apriori guess is that the selected purchasers/leasors are going to win an once-in-a-life-time jackpot!

Recommendation:
The qualified, go and prepare well for the lottery NOW!

睇報學經濟 - Does Solvency Matter?

Paul Krugman, 2008 Nobel Laureate in Economic Science, wrote on Social Security in his NYT weekly column this Monday. He said that Social Security will remain solvent at least until 2030's and therefore it is illegitimate to support any proposal on Social Security cut (ibid).

The article is, of course, more than a strike back to Social Security cut proposer (actually the article is about fiscal austerity). Nevertheless, among the various arguments raised in his article, one issue pops up (and keeps popping up) in my mind, and that is "Does Solvency Matter" or better phrased as "Only Solvency Matter".

Yes, we "know" (I presume I know) Social Security is unlikely to go bankrupt until 2030's, but does that mean we shouldn't cut Social Security?

Again, the answer (for me at least) is "NO, NOT NECESSARY".

Argument 1:
Doesn't ECON101 say (social?) utility is maximized when marginal benefit equals to marginal cost, but not when the action's (net?) utility remains positive (or non-negative)?

Argument 2:
If so, does that mean to cut or not to cut, solvency is "a", but not "the sole", consideration?

Does that mean Mr. Krugman is wrong?

Ans: No, he doesn't, given that

1.
2.
... ...

Dear reader, I intentionally leave the conditions blank and make it a challenge to you to spot them out (hints: under what circumstance(s) does the condition of having non-negative utility an optimal? what happens when average cost equals to marginal cost? under what circumstances will that likely happen?). I also leave it to you to decide if the underlying Krugman philosophy is sound and followable.

To trust or not to trust, that's your own question.

Does Speculation tend to Stabilize Prices?

Ed discussed speculating activities in his blog. I didn't read his paper on speculation (as I don't have the full access to it) , so I cannot comment on its conclusion "禁止摩貨勢必令市價波動幅度增加".

The paper's conclusion may be correct (and its validity is not this post's subject). But, theoretically, speculation does not necessarily tend to stabilize prices! It usually does when we use the term "speculation" for what is "arbitrage" (over time, or across different participating parties, or a mix of both). The problem is that speculation can be postulated, in addition to arbitrage, as buying when the price path is rising and selling when it is falling, in which case price volatility increases.

I suggest the following to interested readers:-

1. section 6.2 of J. Hirshleifer and J. Riley's "The Analytics of Uncertainty and Information"
2. Baumol, W., "Speculation, Profitability, and Stability," Review of Economics and Statistics, 30 (August 1957)
3. Friedman, M., "In Defense of Destabilizing Speculation," reprinted in Friedman, M. The Optimum Quantity of Money and Other Essays.

睇報學經濟﹣鄧監製

無記前金牌監製鄧監製接受壹仔訪問，說無記出品十年如一日，內容對白白痴到不能接受程度。

鄧監製在訪問中指 “外國劇，內地劇，製作模式跟無綫出品，已經完全不同 ... ... 只不過無得揀。收看免費電視，基本上可以話係無選擇。無競爭，管理層自然樂於因循下去 ... ...”， 並說 “流水作業，一樣可以有性格。監製絕對有權按照流水作業方式在有限成本下擺自己的簽名落去。搭景，係咪可以有少少要求，搭靚啲用多了錢，咪慳在其他地方囉。”

鄧監製的出品如何及其對電視台營運的看法是對是錯非本文評論範圍。本文重點乃鄧監製對搭景的意見。經濟學假設人尋最大利潤（maximize profit)，而得最大利潤的均��點(equilibrium position)為每生產要素的邊際利益（marginal product of x)等如其邊際成本(marginal cost of x) - x = 演員、景、編劇 ... ... 等。於是要賺盡，無記劇的生產非「搭靚啲景用多了錢，咪慳在其他地方囉」，而是令每生產要素x的邊際利益等同其邊際成本，即搭景的邊際成本等同其邊際利益，否則無記會少賺了錢 ﹣先在搭景上“虧”，其後在"慳“的地方“賺“少了。

當然，無記朝中有“忠臣”，成功以“海報逼真啲，收視係咪升一點？唔會，唔改 ... ...”制止鄧先生「損害」無記股東之舉。

是股東之幸？是觀眾之不幸？是鄧監製「有志難伸」之大不幸？

鄧監製還是繼續當監製好，盤數還是讓數佬睇吧。比較優勝定律(law of comparative advantage)不是開玩笑的。

Comment on "On Web's Cutting Edge, Anonymity in Name Only"

From Wall Street Journal

On Web's Cutting Edge, Anonymity in Name Only

What (x+1) did was introducing a fact finding algorithm, and, perhaps, suggesting an optimal response function to its client, e.g. Capital One, based on the fact(s) found.

As discussed in yesterday's post, there are two potential problems in the (x+1)/Capital One relationship, namely, the reliability of (x+1)'s "fact" finding service, and, if (x+1) does offer suggestions to Capital One on response, the incentive compatibility between (x+1) and Capital One.

Any (serious) conflict of interests between the two contracting parties can ruin the set-up.

I am interested to know, if possible, a little more about the remuneration clause(s) of the (x+1)/Capital One service contract.

Designing an algorithm shouldn't be considered a violation of privacy. It is the data collecting process which the Authority should focus on.

睇報學經濟﹣陸東說（Andrew Says)

星期日蘋果訪問陸東，陸東說：

“港股現時 13、 14倍 PE（市盈率）” （１）

“萬九至二萬二點，仲要維持一段較長時間”（２）

“與其買滙控，不如買內銀股“（３）

１、２及３皆資訊，但性質不同。

訊息１提供“事實"。受訊者在決定是否相信有關訊息後決定下一步行動（入市or otherwise）。訊息提供人的sincerety 是問題（即陸示有冇呃你, e.g. PE 是否13、14倍）。

訊息２是「機率性訊息」 （而且好大機會包含１的訊息﹣即陸示得悉１後推出２）。Again，受訊者在決定是否相信有關訊息後決定下一步行動。訊息提供人的sincerety 是問題。

３有別於１及２。該訊息（內含２機會甚大）提供下一步行動見意（非純資訊提供）。受訊者要決定是否相信有關recommendation。訊息提供人的sincerety 及其利益與受訊者是否一致是問題。

有趣的問題（之一）是在知道１、２及３時，旁觀者或受訊者能否re-engineer 陸示的決策過程及work back 其對未來state of the world 的預測？如能，在什麼條件之下？

Bob Aumann 的研究（引申）指出在一定條件下，以上問題的答案是 “Yes, we can!”。當然問題不止於此。之後要問的是在己知情況及條件下閣下當下的最佳策略為何。

Aumann 因有關研究（貢獻之一）於2005年諾獎桂冠。

歐民的研究significantly “enhanced our understanding of conflict and cooperation”，並將經濟學研究範圍及approach 推向新世界（唔好搞錯，唔係推俾彤叔）（例如reward structure set-up，contract terms structuring，multi-party conflict resolution... ...）。