Monday, July 25, 2016

New(ish) paper on the Shiller CAPE ratio, by J. Siegel

Published in the May/June issue of the Financial Analysts Journal (vol. 72, no. 3). I'm not sure it it's gated, so you can find a working paper (from 2013!) version here.


Robert Shiller’s cyclically adjusted price–earnings ratio, or CAPE ratio, has served as one of the best forecasting models for long-term future stock returns. But recent forecasts of future equity returns using the CAPE ratio may be overpessimistic because of changes in the computation of GAAP earnings (e.g., “mark-to-market” accounting) that are used in the Shiller CAPE model. When consistent earnings data, such as NIPA (national income and product account) after-tax corporate profits, are substituted for GAAP earnings, the forecasting ability of the CAPE model improves and forecasts of US equity returns increase significantly.

The gist of the paper:

"In this article, I offer an alternative explanation of the elevated CAPE ratio. The nature of the earnings series that is substituted into the CAPE model has not been consistently calculated for the long period over which Shiller has estimated his CAPE equations. Changes in accounting practices since 1990 have depressed reported earnings during economic downturns to a much greater degree than in the earlier years of Shiller’s sample."
"Companies report their earnings in two principal ways: reported earnings (or net income) and operating earnings. Reported earnings are earnings sanctioned by the Financial Accounting Standards Board (FASB), an organization founded in 1973 to establish accounting standards. Those standards—the generally accepted accounting principles, or GAAP—are used to compute the earnings that appear in annual reports and that are filed with government agencies (earnings filed with the IRS may differ from those filed elsewhere). GAAP earnings, which are the basis of the Standard & Poor’s
reported earnings series that Shiller used in computing the CAPE ratio, have undergone significant conceptual changes in recent years. 

A more generous earnings concept is operating earnings, which often exclude such “one-time” events as restructuring charges (expenses associated with a company’s closing a plant or selling a division), investment gains and losses, inventory write-offs, expenses associated with mergers and spinoffs, and depreciation or impairment of “goodwill.” But the term operating earnings is not defined by the FASB, and companies thus have some latitude in interpreting what is and what is not excluded. In certain circumstances, the same charge may be included in the operating earnings of one company and omitted from those of another. Because of these ambiguities, several versions of operating earnings are calculated."
"The definition of reported earnings has undergone substantial changes in the last two decades. In 1993, the FASB issued Statement of Financial Accounting Standards (FAS) No. 115, which stated that securities of financial institutions held for trading or “available for sale” were required to be carried at fair market value. FAS Nos. 142 and 144, issued in 2001, required that any impairments to the value of property, plant, equipment, and other intangibles (e.g., goodwill acquired by purchasing stock above book value) be marked to market.9 These new standards, which required companies to “write down” asset values regardless of whether the asset was sold, were especially severe in economic downturns, when the market prices of assets are depressed. Furthermore, companies were not allowed to write tangible fixed assets back up, even if they recovered from a previous markdown, unless they were sold and recorded as “capital gain” income."
 "A distortion related to the Standard & Poor’s methodology for computing the P/E of an index—what I call the “aggregation bias”—overestimates the effective ratio of the index when a few companies generate large losses, as happened during the financial crisis. S&P adds together the dollar profits and losses of each S&P 500 company, without regard to the weight of each company in the index, to compute the aggregate earnings of the index. This procedure would be correct if each company were a division of the same conglomerate and one wished to determine the P/E of that conglomerate"
"Because of changes in the definition of GAAP earnings, it is important to use a definition of corporate profits that has not changed over time, as in the series computed by the national income economists at the Bureau of Economic Analysis (BEA), which compiles the national income and product accounts (NIPAs)."
"In forecasting future 10-year real stock returns, the highest R squared is achieved by using NIPA profits for specifications of the CAPE regression, with either the price index portfolio or the total return portfolio."

Siegel offers alternative estimates of how over-valued the S&P is, according to each CAPE measure, as well as estimates of future returns. The CAPE that uses the NIPA profit measure produces the lowest over-valuation and the highest expected returns. I'm generally sceptical of such estimates, so I won't go into those details.

What I got from this paper is a reminder that the S&P measure of profits (and perhaps other measures that rely on reported earnings) has changed over time, due to accounting changes, so one has to be careful when using it.

Friday, May 13, 2016

Robots for everything: Teaching assistants

From the Washington Post:

(Thanks to @JustinWolfers for his tweet.)

My thoughts after reading this article:

1. Robots can be more valuable when there is more technology in the environment.  In this example, a chatbot was superior to humans in the context of a massive online open course (MOOC),
"where students often drop out and generally don’t receive the chance to engage with a human instructor. With more human-like interaction, Goel expects online learning could become more appealing to students and lead to better educational outcomes." 
This brings up the possibility of increasing returns to capital, at least up to a point. Here, the internet made MOOC's possible, which in turn made robots more valuable.

2. If it's important to make the interaction with the bot human-like, make your robots more human: program them to respond with a delay, to make mistakes, etc. In this example:
"I had the same doubt last week [whether the TA was a chatbot] because we were getting such speedy responses from TAs."
3. Context may be important for how well customers accept artificial intelligence. In this example, the users were taking an AI-related class, so it's natural the chatbot would be welcome:
" 'A really fun thing in this class has been once students knew about Jill they were so motivated, so engaged. I’ve never seen this kind of motivation and engagement,' Goel said. 'What a beautiful way of teaching artificial intelligence.' "

Thursday, May 5, 2016

"I still hold out some hope": James Bullard on the U.S. becoming Japan

Refreshingly outspoken James Bullard, from the St. Louis Fed, gives an interview to the New York Times. Some great quotes (emphasis mine):

"The general rule of thumb within the Fed is that labor market data trumps G.D.P. data.
"Once major central banks hit the zero lower bound, the key issue was whether central banks would be able to keep inflation expectations consistent with inflation targets.
"I think the Brexit vote, there are a couple of aspects of this that make it much less of an international macroeconomic event. It’s a scheduled event; you can track which way the vote is going to go by looking at polling; and it’s a long-term strategic vote on the part of the U.K. The day after Brexit — even if they vote to leave — nothing would actually change in terms of the trade arrangements. Those would continue for at least two years." 
"The norm in central banking, away from the zero bound, is to say, “We have set the policy rate exactly where we think it should be for today, given everything that’s going on in the economy, and in the future we’ll look at the data.” You didn’t do this kind of dot-plot thing. [...] I’ve wondered if we should get back to something that’s more akin to that. We don’t want to give unintentional commitments." 
"I’ve actually argued that unconventional policy works reasonably well. But it’s far less clear how it works, or how effective it is." 
"I’ve always been worried that the long run here is the Japanese outcome. I still hold out hope that that’s not the case, but I am worried about it, and it’s been going on for a very long time. If you talk to people in Tokyo, they say, 'Well, we’ve been through this and tried all these things, and you guys are just following us.' I hope that’s not exactly true. [...] I still hold out some hope.
"I’m not as big an advocate of fiscal policy as some other people. It’s very hard to do very much on a business cycle time scale, given the fact that you’ve got to work with Congress.
"At some point something will happen and we’ll be back in recession, and by almost any reckoning we will not have much that we can do in the way of lowering our policy rate."

Thursday, April 28, 2016

Q1 GDP nowcasts

Atlanta Fed's nowcast of Q1 GDP (0.6%) was really close to preliminary estimate announced today by BEA (0.5%). Atlanta's was much closer than the New York Fed's nowcast (0.8%). The question is which of the two nowcasts will be closer to the final estimate, after revisions (available in late June).

Gavyn Davies and Juan Antolín-Díaz explain why these two nowcasts can differ so much from each other, and from their own nowcast at Fulcrum Asset Management. The Atlanta Fed's approach consists of "bean counting," i.e. they mimick the way the BEA calculates GDP by aggregating monthly data as they are released. The New York Fed's and Fulcrum's methodologies are both dynamic factor models, which extract a "common factor" from multiple time series (not only those used by BEA to estimate GDP). This underlying growth can then be scaled to match certain properties of the GDP time series (as the New York Fed does), or not be scaled (which is the approach they prefer at Fulcrum).

Monday, December 21, 2015

Brief comments on the elections in Spain

A little closer to Italy. The fragmentation of the vote was the expected and actual result of yesterday's general election. Incumbent Partido Popular (PP - moderate right) won the most votes, but came far short of the majority it had obtained in 2011.

The main other party, Partido Socialista (PSOE - moderate left) might be able to form a government coalition with the new Podemos (far left), but they will still need the support of a motley crew of smaller parties in order to reach the 176-seat majority. Another possibility is a core coalition of PSOE with Ciudadanos (C's - center), plus somebody else. But that seems even less likely, as Ciudadanos promised during the campaign that they wouldn't join forces with either Podemos or nationalist parties. A grand coalition of PP and PSOE, à la Germany, would be unpalatable for PSOE - and its demise as the leading party on the left.

Source: El País.

After the first vote to form a government, which the incumbent prime minister will presumably lose, the parties have two months to form a government coalition before they must call fresh elections.

Any coalition will be uneasy and precarious. Partido Popular has made bitter enemies during its four years in government. It's been an acrimonious election campaign. The programs differ wildly across parties. Whatever government they form, I think significant advances on important legislation are unlikely over the next four years (labor market, changes to the Constitution, education, Catalonia's independence referendum).

Another important (and surprising to me) result was the decline of nationalist parties in Catalunya and Euskadi, and the rise, in those same regions, of the new left-wing party, Podemos. As a pro-referendum force, Podemos could actually do more for the independence cause than either the Catalan or Basque nationalists on their own. Yesterday's elections gave Podemos 69 seats, more than all the nationalist parties combined ever got in any general election. It's still doubtful, however, whether Podemos will be part of the new government. Moreover, those 69 seats won't belong to a single parliamentary group, as Podemos is itself an umbrella brand that includes a cluster of regional, left-wing parties with their own agendas.

Messy. Bumpy. And thoroughly entertaining.

Election results, town by town
Can you spot which two regions are different?
Source: El País.

Wednesday, December 16, 2015

The hawkish shift of monetary policy recommendations

A recent poll by Chicago Booth (thanks, Tyler!) asked a panel of economists in early December to agree or disagree with the following statement
The Fed should raise its target interest rate when it meets in mid-December.
The respondents (all of whom are "senior faculty at the most elite research universities in the United States") were largely in favor of a hike. Forty-eight percent either agreed or strongly agreed; and 19% disagreed. Nobody disagreed strongly.

Digging through previous questions to the same panel, I found a puzzling result from early April. The question then was:
The Fed should wait until its preferred measure of inflation (Core PCE) is clearly rising — and not just forecast to rise — before it begins hiking interest rates.
Here's how the results of the two polls compare:



The hawkish swing is dramatic. In April 40% of respondents thought the Fed should wait for core inflation to rise. In December, even though core inflation had not risen at all, only 19% didn't think the Fed should raise immediately. What's going on? (Remember, both statements were about what respondents thought the Fed should do, not predictions of what the Fed would do.)

1) Unemployment went down. But I find this hard to believe, because the unemployment rate fell just 0.5% between March and November, and the consensus forecast in March was already that unemployment would go down.

2) Despite the wording of the April question, respondents did have in mind forecast inflation, not just actual inflation. But forecasts of core inflation barely changed between April and December. If anything, medium-term forecasts went down slightly.

3) There is a third (fourth, fifth?) variable in the respondents' mental Taylor rule. However, this additional variable would have to produce a hawkish leaning by December. Output didn't grow particularly fast, in the U.S. or globally. Perhaps a heightened awareness that low interest rates are destabilizing the global financial system? I'm not convinced because, if anything, the consensus is probably that the system is more fragile now than in April. An interest rate hike, at the wrong time, could trigger the crisis.

4) Respondents are concerned about the Fed's credibility. The Fed had been beating the hiking drum from October through December. Doing nothing in December would have undermined the effectiveness of future Fed communications.

5) The respondents, by December, believed that the December hike was largely testimonial, a mere assertion that the Fed is still "in charge," and the hike is to be followed by a gentle tightening cycle. This was not the consensus in April yet.

6) The poll statements are poorly written, because they fail to account for a behavioral bias that makes respondents more likely to agree than disagree with whatever statement they face.

7) Respondents are unconsciously conflating the normative statement with predictions of what the Fed will do. (The consensus prediction shifted dramatically between April and December, mostly because the Fed itself gave strong guidance of an interest rate hike in October and November.)

8) Tyler is right, and economists don't know what they're talking about. Moreover, their ignorant self-confidence produces time-varying biases.

I think a combination of #3, #4, and #5 is most likely, but (as a respondent to opinion polls myself) I can't completely dismiss behavioral biases of the respondents.

Wednesday, October 28, 2015

Talking worth hearing: Accounting for profits

Stan Pignal, Patrick Foulis, and Philip Coggan (all three from The Economist), talk about how managers find (mostly) legal ways to puff up corporate earnings.

~12 minutes


Thursday, October 22, 2015

Talking worth hearing: China

Tyler Cowen talks about the rise and fall of the Chinese economy: how they grew so fast, and why they're in trouble now.

~12 minutes

I like Tyler's insight that decades of high growth distorted the assessment of the risk/return of new investment projects, which is distinct (but compatible) from a decline in the marginal productivity of capital.