Dear friends,
I thought WallStreetRanter's note on the current price/book valuations in e.m. stocks (i.e., really low) and the historic pattern of returns (i.e. really high) after hitting such valuations was striking. GMO today released their latest seven-year asset class real return forecast. They project emerging market equities to be the highest returning asset class at 6.8% real (i.e., after inflation) returns and emerging market bonds to be the highest returning bond class (at 2.5% real).
Other asset classes:
3.1% - U.S. high quality stocks
2.4% - International stock cap stocks
2.1% - International small cap stocks
0.2% - U.S. aggregate bonds
0.2% - Inflation-linked bonds
(0.4%) - Cash
(2.0%) - International bonds
(2.1%) - U.S. large cap stocks
(3.5%) - U.S. small cap stocks
In a related note, I added to my Seafarer holdings this week.
For what it's worth,
David
Comments
Inker: Well, we have not actually been acting as if we believed that 7% forecast. That 7% forecast is assuming everything goes back to normal, and one of the things we've been concerned about and that we have written about is the problem of China... As a result, we have been more cautious on emerging markets than our forecast would suggest, and we still are. Emerging markets have some real problems, and we are trying to figure out exactly how they will play out. We still think you've got to own some, because they are pretty cheap even if they do have some problems. And certainly having underperformed the U.S. by 27%, 28% so far this year, they are getting more attractive. I'm not sure they are quite as attractive as the forecast, which is assuming we get to go back to normal, when it isn't entirely clear what normal is for these economies.
We could say it's going to be 7%, but it's going to be a bumpy ride, so hold on... The biggest concern we have is effectively that 7% return is assuming that more or less all of the very strong growth emerging markets have had for the last decade was secular growth. If some portion of it was instead this cyclic goal issue of selling into a bubble in China and when that bubble bursts, the sustainable level of sales and earnings and assets will all be lower, then it's not just that you get a bumpy ride, but you're not going to get the 7%. And that's what worries us more.
We're perfectly happy to take a volatile 7%, and (we) may well start buying some more over the next couple of months, but the big concern is we don't quite believe it's as cheap as our standard analysis makes it look.
Are the more traditional EM funds passe? Absolutely not, but we should probably not expect the outlying kinds of returns these have had in past decades.
Of the consumer-oriented EM/FM funds I follow, WAFMX looks like the best valuation; it, THDIX and ECON are at relatively high P/B's, but the latter two are also up there in P/E and P/CF, while WAFMX is still pretty reasonable by those two measures.
Emerging Markets Dividend Growth ETF (DGRE)
http://www.wisdomtree.com/resource-library/pdf/materials/DGRE/WisdomTree-Case-For-DGRE-1448.pdf
Does anyone bother to keep score of the forecasting follies game?
Okay, the respected GMO fund firm just released a worldwide forecast of expected future real returns for various investment classes. That’s comfort food for those planning to commit to the projected high return areas anyway. I hope the report did not influence those who were not so inclined to join the stampeding herd. That’s how bubbles are initiated.
I like the GMO organization, and I especially respect Jeremy Grantham. They’re fine folks and he is a true knowledgeable gentleman. But has anyone scored the accuracy and reliability of their far ranging forecasts? How good are they in the forecasting business?
Forecasters of all kinds, in all sizes, in all disciplines for all times have prospered forever. Historically, that prosperity is derived from their fees, not from their prescient prognostication abilities.
Probably the most extensive study of forecasting accuracy was conducted by professor Phil Tetlock over a 20 year period consulting roughly 300 experts (most often in the political arena). Over 70,000 forecasts were evaluated and scored. On average, these wizards scored just over 50 % correct. A “no change” judgment would have done about as well. A simple linear extrapolation algorithm would have bested the wiz-kids in many instances. From this work, Tetlock coined the Fox and Hedgehog grouping of experts. He trusted the Foxes of the world.
Way back in a1932 paper, Alfred Crowles 3rd asked the mutual fund management question “Can Forecasters Forecast? His research data answered a definite “No”. That early research finding has been echoed time and time again by an endless stream of academic and industry research work. The WSJ and Barron’s assess their stock picking genius frequently. Sometimes they exceed dart throwing monkeys: sometimes they don’t. In all cases, the margin of victory is thin.
Reacting to market guru forecasts is hazardous to your end wealth.
There is no doubt that the GMO family of funds is a quality operation. They own a respectable overall performance record. By my stars, they tend to underperform slightly in a bull market environment, but demonstrate their mettle with superior rewards when the marketplace is in serious turmoil. That’s an attractive and worthy attribute.
But how good are they at forecasting future real returns around the globe? That’s a daunting challenge that they must accept when they publish predictions. It’s fair that GMO should be prepared to be rated equally on all its forecasts; both its good ones and its bad ones. But that doesn’t ever seem to happen.
Typically, forecasters get to tout their victories and are permitted to conveniently forget their failures. I have a higher standard. Before I trust a forecaster, I need access to his entire forecasting historical record. Of course I am never granted that access, so I basically distrust all forecasts and do not react to them.
But just one moment please. Jeremy Grantham does have a verifiable prediction record. It was generated over the years by the CXO Advisory Group. It is contained in their Gurus section. Here is the Link:
http://www.cxoadvisory.com/gurus/
In reviewing the listing, Jeremy Grantham made 37 evaluated forecasts and received a 47,5 % accuracy grade. Not too bad; not too good. A lot like tossing a coin. Comparatively, he scored in the middle of the total guru distribution population. I’m singularly not impressed.
A seven year forecast is a long stretch given that forecasting just 6 months ahead is tough. Forecast accuracy degrades as the time dimension increases. I hope you guys did not invest in emerging markets based solely on the GMO forecast. That’s a risky adventure, maybe even a gamble.
I stay away from forecasts, indeed, very far way. They usually are sheer nonsense, equivalent to financial pornography.
Best Wishes.
http://www.forbes.com/sites/kenrapoza/2013/08/09/for-emerging-markets-better-luck-next-year/
In general, GMO seems to get the rank-ordering mostly right (their highest projected asset class is likely going to end up #1) and the order of magnitude mostly right (if they say it'll be modest negative, it's likely to be negative. GMO does give error ranges for each projection, which some being really substantial.
For what it's worth,
David
Hi Guys,
Thank you both for the Duke and CXO references. I was totally unaware of their existence. I was able to access both sources and have quickly scanned them. Good stuff.
My immediate reaction, based on their bottom-line conclusions was a definite wow. The GMO team do superior forecasting work. Perhaps they indeed are the exception that prove the more common dismal forecasting rule.
Upon a deeper, reflective reading of the documentation, I feel the need to curb my original enthusiasm, at least just a little.
The Duke University brief study showed overall prediction correlation factors in the 0.78 to 0.94 range. That’s pretty impressive, but certainly not perfect. Although the GMO predictions got the direction right for all 3 bond cases (that’s relatively easy), it failed directionally in 3 of the 5 more challenging equity cases examined. Although the return magnitudes were only off by a percent or two, the percentage errors were in the double digit ranges. Not bad, but not exceptional.
Within the text of the Duke report, the researcher observes that since the highest ranked class did not always dominate the annual returns the results are “….. suggesting limits to the trust one should put in the GMO predictions as a tool for short-term prediction”. That’s not an unconstrained endorsement of the GMO methodology.
As John Bogle demonstrated in the first edition of his “Common Sense on Mutual Funds” book, longer term performance predictions are far more reliable than shorter-term projections. The longer-term forecasts benefit from complete market cycles and from the ubiquitous regression-to-the-mean tendency of the marketplace. Bogle shows the consistency of equity returns over many decades of 10-year increments.
The CXO acid out-of-sample test examined the GMO 10-year forecast made at the end of 1999 against the actual market results for that decade, ending in 2010. The GMO group did an excellent job at predicting class rankings. They demonstrated skill in that measurement well beyond a random chance success. Once again the disparity between actual and predicted returns was double digit on a percentage basis.
CXO concludes that “building an investment strategy around these long-term forecasts seems problematic”,
CXO Advisory Group also highlighted the forecasting accuracy difficulty of more volatile short-term performance estimates against the less violent aspects of longer-term predictions. In the marketplace, time is a smoothing operation. The market data illustrates that trend.
Given the rather modest return forecasts typically featured in the GMO quarterly predictions, Phil Tetlock would probably classify the GMO team as Foxes. They are likely to produce more accurate, but less dramatic predictions. Studies suggest that Foxes have a higher prediction accuracy score over Hedgehogs.
I love this stuff. I wonder if there have been more recent updates of studies like those completed at Duke and at CXO? Both Duke and CXO do commendable, investor exploitable work.
Thanks again Guys. You made my day. Just a little forecast scoring is actually done. That’s goodness.
Best Wishes.
http://www.marketwatch.com/story/emerging-market-fears-hide-enticing-valuations-2013-08-16