Roubini’s Real Track Record
Posted by Eddy Elfenbein on October 11th, 2010 at 4:13 pm

I’m glad to see Charles Gasparino’s column on Nouriel Roubini’s overhyped reputation:

A closer inspection of Roubini’s record shows that while he was predicting doom and gloom for the US in 2004, his initial call had nothing to do with a runaway housing bubble.

Rather he argued that the Bush Administration was racking up massive deficits to foreign investors, namely the Chinese, and that the Chinese would scale back on their purchases of US debt, causing interest rates to spike and the dollar to decline in value, resulting in “financial trainwrecks for the US economy in a matter of a couple of years.”

Sounds good, but the problem with the theory is that it didn’t happen.

While I’m as worried as anyone else about the Chinese financing US domestic spending, it should be noted that there’s little sign that they’re about to stop, even with the Obama administration making President Bush look like a deficit hawk.

Based on my research, it wasn’t until about August 2006 that Roubini began talking about a housing crisis, and he was hardly alone. Several economists and investors, from John Paulson to Stan Druckenmiller and around this time Goldman Sachs, were also predicting the housing decline.

Roubini, for his part, wasn’t available for comment, but a spokesman said in a statement that he has “built a strong and rapidly growing business,” and has more than “1,000 institutional clients.”

Maybe so, but how do they feel about his call on the price of gold? Last year he predicted that the rising price of gold was in fact a bubble, just like the housing one a few years earlier, and like housing, it would burst as well. But as we all know gold prices remain strong.

For the record, I think Roubini is a very smart guy and well worth listening to. But I don’t believe he predicted the financial crisis.

There are currently 837,000 Google matches for the search “predicted the financial crisis.” The top one, as it turns out, was written by me.

If so many people saw it coming, it’s a wonder how it happened.

The Federal Reserve’s Educational Links
Posted by Eddy Elfenbein on October 11th, 2010 at 3:55 pm

The Federal Reserve has a number of education resources, often in the form of online games. Here’s a quick rundown:

The Federal Reserve’s Kid’s Page

Great Economists Treasure Hunt

Welcome to FedVille

Escape from Barter Islands

Fed Card Currency Trivia Game

Fed Chairman Game

Peanuts & Crackerjacks, the Federal Reserve Bank of Boston’s interactive baseball game

Finally, are you looking for a Fed membership form? I got ya right here.

Scary Jobs Chart
Posted by Eddy Elfenbein on October 11th, 2010 at 1:55 pm

The number of Americans who have been unemployed 27 weeks or longer:

No Empirical Support for the 200-DMA?
Posted by Eddy Elfenbein on October 11th, 2010 at 1:13 pm

This is from the Q&A section of the Fama/French Forum:

Some researchers argue that a market timing strategy based on buy/sell signals generated by a 50- or 200-day moving average offers a more appealing combination of risk and return than a buy-and-hold approach. What is your view?

EFF/KRF: An ancient tale with no empirical support.

That’s simply not true. This is what I wrote about the 200-DMA last year:

One of the quick-and-dirty tools used by technical analysts is to see where a stock or index is compared with its average price over the past 200 days. This is an easy way to get a read of a stock’s momentum.

Yesterday was a big day for the 200DMA world. The S&P 500 closed above its 200DMA for the first time since December 26, 2007. That closed out the index’s longest run below its 200DMA according to my records which go back to 1932.

That streak, however, is still well short of the longest run above the 200DMA which ran from November 1953 all the way to May 1956. Since the index has gone up over time, the “above” streaks tend to be longer than the “below” streaks.

On November 20, 2008, the S&P was a stunning 39.6% below its 200DMA. That’s the biggest discount on my records. The only thing that comes close is the reading from this past March.

So does the 200DMA work? The evidence suggests that it’s a pretty good indicator of future price performance. When the S&P 500 has been below the 200DMA, it’s dropped a total of about 20% over the equivalent of 27 years. In other words, the S&P 500 has been below its 200DMA about one-third of the time.

Historically, the best time to invest has been when the S&P is less than 1.7% below the 200DMA.

When the index is above the 200DMA, well, then everything looks much brighter. All of the market’s gains and then some have happened when we’re above the 200DMA which occurs about two-thirds of the time.

The market seems to like nearly every point of being above the 200DMA. Danger only clicks in when the S&P 500 is over 17.5% above the 200DMA which is a very high reading.

This issue isn’t whether the 200-DMA works or not. It’s a dumb rule, but it reveals an important truth about investing: the market likes trends. If the market is going in one direction, it has a much better than average chance of staying in that direction.

Happy Columbus Day
Posted by Eddy Elfenbein on October 11th, 2010 at 11:31 am

Columbus Day is unusual for Wall Street because the bond market is closed yet the stock market is open. I have absolutely no idea why. I believe the Columbus was financed by the original Sovereign Wealth Fund, with actual sovereigns!

Three months ago, I said Intel (INTC) would beat its earnings report and I was right. Then last month, I said Bed Bath & Beyond (BBBY) could earn as much as 70 cents per share. That was a gutsy call since it was well above Wall Street’s expectations. Nevertheless, I was right again.

Now Intel is due to report earnings again tomorrow. The Street’s current estimate is for 50 cents per share, and this time I’m not expecting much of an earnings beat, if any at all. Fifty cents per share sounds about right.

The stock that I think is a good candidate for an earnings beat is JPMorgan Chase (JPM). I hope to post more of my thoughts on that this week. The bank will report earnings on Wednesday. Except for Intel, the earnings reports for the rest of our Buy List won’t start coming in until next week.

Morning News: October 11, 2010
Posted by Eddy Elfenbein on October 11th, 2010 at 7:26 am

Treasury 2-, 5-Year Yields Decline to Record Lows on Outlook for Fed Buys

Is Sysco the Perfect Stock?

Richard Bernstein to Manage Stock Mutual Fund for Eaton Vance

Dow 11,000…Again

Everyone into the Pool: How to Invest in Twitter

Happy Fifth Blogiversary to Abnormal Returns

European Stocks Edge Higher Amid QE Hopes

Wage China Currency War With Light Armor, Investors Say

Foreclosure Freeze May Sideline U.S. Homebuyers as Legal Worry Cuts Sales

Why is Norm Van Brocklin still a record holder?

What’s the Future for Gold?
Posted by Eddy Elfenbein on October 10th, 2010 at 10:44 pm

I took the current TIPs yield curve and plugged it into my Gold Model to see what the future may hold for gold:

Now let me add several major caveats.

This is not in any way my forecast for where gold will go. The model I made is just that, a model. Or more specifically, it’s a model of how a better model might look.

As to the specific constants I used (eight for leverage and 2% for equilibrium), those are just working estimates. If those numbers are changed even slightly, the chart above looks very different.

Of course, the numbers can also lead to an entirely different conclusion: that TIPs are far too high. (A real return of 0% for six years? No thank you!)

We should also bear in mind that estimates for the future based on what we know today are notoriously poor. In 1999, the TIP that was maturing in 2002 was showing a real yield of 3.5% to 4%.

The real takeaway is that the market thinks real rates will remained subdued for a long time. It won’t be another 10 years until real rates get back to normal.

Implied Markets — House Elections
Posted by Eddy Elfenbein on October 10th, 2010 at 8:39 pm

One of the topics in finance I find most fascinating is implied prices. This is when we take the prices of two or more assets and use them to find an “implied” price for another asset.

For example, that’s what the VIX is all about—implied volatility. In the options pricing model, volatility is one of the variables. So what we can do is take the current market price of an index option and work backward to see what number the market has given to the volatility variable.

I’ve noticed that we can also find some implied prices in futures markets on current events listed at Intrade. (Note: This is not — repeat, not — a political post. It’s about implied markets.)

Here’s an example. The current price for the Republicans to win control of Congress is 78.2. That’s a gain of 40 seats. There’s another contract for the Republicans to gain 50 seats. The latest price for that contract is 52.0. (Second note: I don’t take these markets too seriously. I simply see them as entertainment.)

A 78.2% probability translates to +0.779 standard deviations above the mean (that’s NORMSINV in Excel). A 52% probability translates to +0.050 standard deviations above the mean. The difference between those must be 10 seats, so one full implied standard deviation works out to 13.7 seats.

From there we can see that the mean is a gain of 50.7 seats. With the GOP currently holding 178 seats, the market therefore currently expects the GOP to hold 228.7 seats after the election.

I downloaded the historical numbers and here’s what the market’s implied outcome is for the election. The black line is 218 which is a majority. The blue line is the implied mean for the GOP. The red line is one standard deviation above the mean and the green line is one standard deviation below.

Remembrance of Stocks Past
Posted by Eddy Elfenbein on October 10th, 2010 at 3:57 pm

One of my rules of investing is: “Never worry about what stocks do after you’ve sold them.” The stocks aren’t thinking of you, so why should you think of them? That’s a wise rule. Nevertheless, I’m ignoring it for the moment.

Each year, I add and delete five stocks to my 20-stock Buy List. As it turns out, the stocks I removed from the Buy List at the beginning of this year have done far better than their replacements.

Here are the five new stocks:

Here are the five stocks I cut:

Even though I limit myself to changing the portfolio by 25% just once per year, even that was too much. If I had made no changes at all, the Buy List would be up 11.59% instead of 6.97%.

OK, I’m going back to enforcing my rule. If anyone needs me, I’ll be pretending CTSH isn’t up 41% YTD.

Gold Model — Reader Response
Posted by Eddy Elfenbein on October 9th, 2010 at 4:50 pm

Here’s a thoughtful email I received in response to my gold model:

I read your piece on your gold model and thought it was very good – really dead on. I think, however, you have the relationship somewhat backwards on two counts.

First, the dollar derives its value from its relationship to gold. So instead of the dollar price of an ounce of gold, it should be thought of as the gold price of a single dollar. (For instance, the current of gold price of single US dollar is presently about 1/1345th of an ounce of gold)

Second, on this basis, the relationship is somewhat clearer: when the purchasing power of an ounce of gold rises (i.e., when an ounce of gold commands more dollars) interest rates will be low; and, when the purchasing power of a an ounce of gold falls (i.e., when gold commands fewer dollars the interest rate will be high.

In the days of the gold standard, this meant when prices, denominated in gold backed money, fell, interest rates fell – for instance, during periods of depressions. This would also be the period when the purchasing power of gold backed money was at its highest.

When prices, denominated in gold backed money rose, interest rates also rose – for instances during periods of rapid expansion. This would have been the period when the purchasing power of gold backed money was at its lowest.

Under the gold standard, the purchasing power of money fell during expansions, or what is the same thing, prices rose. And, interest rates rose with prices. The purchasing power of money rose during depressions, or what is the same thing, prices fell. And interest rates fell with prices.

So, what you have shown is that the Gibson Paradox continues to work even in the absence of gold backed money. It did not disappear. It expresses itself through national currencies and thus reveal the underlying state of the real economy through this operation.

When the “dollar price” of gold is rising, it is likely that we are in a depression in the real economy. Hence, real interest rates should be at their lowest. And, when the “dollar price” of gold is falling, it is likely we are in a period of expansion. Hence, interest rates should be at their highest.

Both the movement of gold prices and interest rates are determined by the expansion or contraction of the real economy, and by an inverse relation between them. When the real economy is expanding, gold prices will fall and interest rates will be higher. When the real economy is contracting, gold prices will rise, and interest rates will be lower.

What rising gold prices are telling us today, and for the last decade, is that we are in a depression in the real economy – one that has lasted about a decade at present, and shows no signs of ending as yet. Layered over this real economy depression has been two monetary recessions – the first in 2001, and the present one; with, one period of monetary expansion between them.