Month: July 2009

It Takes Two to Tango

It Takes Two to Tango

For every buyer there is a seller.? For every debit, there is a credit.

People often accept naive views of how the market works, perhaps considering how their own life seems to work, and not considering? the other side of the trade.? As an example, aside from laziness, why do market observers report a day where the market goes down on no significant news as “profit taking,” or grab at some lame smaller story which couldn’t explain the decline?? For every seller, there is a buyer.? No money went into or out of the market unless there was a new IPO, rights offering, company sale for cash, buyback, cash dividend, etc.? People don’t run away from or run to the market; only the terms of the tradew change at the margin.

The same applies to current account deficits.? They have to get funded from somewhere, and on unfavorable terms to the lender if the borrower happens to be the world’s reserve currency.

Thus, when I consider arguments over whether America is to blame for its profligate ways, or whether those that funded the deficits are to blame, I simply say that the books have to balance.? Neither is to blame; both are to blame.

It is not as if China has free capital markets.? Given their neomercantilism (uneconomic export promotion), they had to find places where their exports would be accepted.? The answer was the US.? After that, what do their banks do with excess dollars?? They buy fixed income dollar assets, which they foolishly think will preserve value until they need to liquidate the assets for goods or services of some sort.

That recycling of the current account deficit forced rates lower in the US while the Fed was tightening.? For the Fed to have fought that influence, they should have tightened more rapidly, compared to the plodding 1/4% each FOMC meeting.? How often have mortgage interest rates fallen while the Fed is tightening?? Not often, which is why the Fed was impotent during the last tightening cycle.? It is also why the blows hitting the global economy have fallen more lightly on the US.? To the extent that foreigners buy our bonds denominated in dollars, that transfers a part of the pain to them.? Thanks, but you could have avoided our pain had you opened your markets to our goods and services.

There are many efforts in play to try to replace the dollar.? Most if not all will fail.? At present, the US is politically secure in ways the other large currencies are not, and many invest in the US not to preserve full value, but to preserve most of the value, whatever that may be.

As with many things in life — it takes two to tango.? Blame is infrequently singular.? Both the US and China should own up to their shares of the current problems.? Then, maybe, solutions could be found.

Efficient Markets as a Limiting Concept; There are Conceptual Limits to Efficient Markets

Efficient Markets as a Limiting Concept; There are Conceptual Limits to Efficient Markets

I do and don’t believe in the efficient markets hypothesis [EMH].? I do believe in the adaptive markets hypothesis [AMH].?? The efficient markets hypothesis posits that:

  • Past price-related information can’t be used to obtain better-than-average returns. (weak form of the EMH — cuts against technicians)
  • Past and present public information can’t be used to obtain better-than-average returns. (semi-strong form of the EMH — cuts against fundamental analysis)
  • Public and private information can’t be used to obtain better-than-average returns. (strong form of the EMH — believed by few)

In practice, the academic community holds to the semi-strong? form, while the investment community holds to the weak form.? One thing is certain: the market is dominated by large institutions, and the market on the whole, less fees, cannot beat the returns of the market on the whole.

Part of the problem with the EMH is that with respect to the market as a whole, of course it is true.? The real question is whether any particular strategy covering a small portion of the assets of the market can consistently beat the returns of the market on the whole.? I believe the answer to that question is yes.

An implicit assumption of the EMH is that research costs are free.? They are not free.? Also, it implicitly assumes that a dominant number of investors understand what information drives the markets.? Both assumptions are not true — even in the most clever firms, there is information that is missed, and research costs are expensive, and not always rewarded.

But the effort to earn above-average returns forces the market closer to the EMH.? When the competition is tough, finding excess returns is hard.? This makes it a limiting concept.? We never get there, but effort to find above-average returns gets us closer to that ideal.? Conversely, when many decide to index, those who do not index have a better chance at earning above normal returns, because there is a large chunk of naive capital in the market seeking average returns with certainty.

I want average people to use index funds for many reasons:

  • It lowers their costs.
  • It is tax-efficient.
  • Most people aren’t very good at picking equity managers.? They go for the manager who is hot, in the style that is hot, rather than one that did better in the past, and is in a cold spell now.? They go for large fund groups that spread their research over large asset bases, diluting whatever skill they might have.?? The best managers are the smaller specialists running their own funds, and who eat their own cooking.? They are also inconvenient to use.
  • It improves conditions for the remaining active managers.

I also want them to buy-and-hold (dirty words) because they aren’t very good at market timing, and also have enough in safe assets to lower the downside of returns to a level that does not panic them.? Most people are bad at most investment decision-making.? Better to hand it off to those who don’t panic or get greedy, than to be a part of those who buy into tops or sell into bottoms.

On the AMH, quoting from another piece of mine of the topic:

The adaptive markets hypothesis says that all of the market inefficiencies exist in a tension with the efficient markets, and that market players make the market more efficient by looking for the inefficiencies, and profiting from them until they disappear, or atleast, until they get so small that it?s not worth the search costs any more.

And so it is for those of us who are active managers.? We have a twofold task:

  • Base our strategies in areas that are unlikely to be overfished for long — e.g., low valuation, positive momentum, and earnings quality.
  • Dip into areas that are temporarily out of favor, whether those are industries, countries, or odd risk factors.? (Odd risk factors: occasionally certain factors in the markets are poison, and even the slightest taint marks a security off-limits, even though those that are barely affected are fine.? My example would be Enron-like structures 2001-2002.? Few would buy the stocks or bonds of companies that had them, even though those structures were not large enough to impair the company, as they did with Enron.? We bought the bonds of a Dominion subsidiary with abandon, because we knew the covenants the bonds had would not kill Dominion, and we had extra value as a result.? What killed Enron benefited us, indirectly.)

To active managers then, I warn: watch how your main strategy goes in and out of favor.? It happens to all of us.? Add to your main strategy most when it is out of favor, and add to whatever alternative you have when your main strategy is running hot.

To average investors, then, I advise: if you adjust frequently, add to your winners and prune your losers.? If you adjust infrequntly (once a year or less), prune your winners and add to your losers.? In the short run, momentum persists, in the longer-term, it mean reverts.

Know yourself.? If you are prone to panic and fear with investments, better to hand the job off to someone competent who will be dispassionate.? If you have conquered those emotions, you can potentially do better yourself in investing.? But ask yourself what your sustainable competitive advantage is in investing.? If you don’t have one, better to index.

Correlation Does Not Imply Causation; A Study of Sector Correlations

Correlation Does Not Imply Causation; A Study of Sector Correlations

There was an interesting post on Bloomberg regarding asset class correlations, and a lot of blogs wrote about it, including Abnormal Returns, which did a nice summary, and expanded the argument to university endowments.? Part of the issue here is that under conditions of stress, assets separate into two simple categories — safe and risky.? To what degree can an asset be turned into cash at anything near its fair value under stressed conditions?

I ran into something similar back in 2006, so I wrote this CC post:


David Merkel
Make the Money Sweat, Man! We Got Retirements to Fund, and Little Time to do it!
3/28/2006 10:23 AM EST

What prompts this post was a bit of research from the estimable Richard Bernstein of Merrill Lynch, where he showed how correlations of returns in risky asset classes have risen over the past six years. (Get your hands on this one if you can.) Commodities, International Stocks, Hedge Funds, and Small Cap Stocks have become more correlated with US Large Cap Stocks over the past five years. With the exception of commodities, the 5-year correlations are over 90%. I would add in other asset classes as well: credit default, emerging markets, junk bonds, low-quality stocks, the toxic waste of Asset- and Mortgage-backed securities, and private equity. Also, all sectors inside the S&P 500 have become more correlated to the S&P 500, with the exception of consumer staples.

In my opinion, this is due to the flood of liquidity seeking high stable returns, which is in turn driven partially by the need to fund the retirements of the baby boomers, and by modern portfolio theory with its mistaken view of risk as variability, rather than probability of loss, and the likely severity thereof. Also, the asset allocators use “brain dead” models that for the most part view the past as prologue, and for the most part project future returns as “the present, but not so much.” Works fine in the middle of a liquidity wave, but lousy at the turning points.

Taking risk to get stable returns is a crowded trade. Asset-specific risk may be lower today in a Modern Portfolio Theory sense. Return variability is low; implied volatilities are for the most part low. But in my opinion, the lack of volatility is hiding an increase in systemic risk. When risky assets have a bad time, they may behave badly as a group.

The only uncorrelated classes at present are cash and bonds (the higher quality the better). If you want diversification in this market, remember fixed income and cash. Oh, and as an aside, think of Municipal bonds, because they are the only fixed income asset class that the flood of foreign liquidity hasn’t touched.

Don’t make aggressive moves rapidly, but my advice is to position your portfolios more conservatively within your risk tolerance.

Position: none

Hmm… in early 2006, we were considerably in advance of the peak that would come in late 2007, but considerably above where we are now.? In my opinion, given my longer timeframe, a good call.

But maybe correlations might rise during times when everyone is anxious to buy risky assets, not just when the want to throw them away.? Do asset correlations rise near peaks for risky assets?

My model on correlations relies on the major industry sectors of the S&P 500:

  • Consumer Discretionary
  • Consumer Staples
  • Energy
  • Financials
  • Health Care
  • Industrials
  • Information Tech
  • Materials
  • Telecom Services
  • Utilities

If lots of money is getting thrown at stocks, won’t the correlations between sectors rise?? And if so, won’t future returns be low or negative?

Here is a graph showing the price return on the S&P 500 over the next 60 days as a function of the average sector correlation over the last 60 days:

Not much of a relationship, huh?? 1% R-squared.? And it goes the wrong way — high correlations very weakly favor higher returns.

But what if we do a regression where future S&P 500 returns are regressed on past S&P 500 returns and average sector correlations?

Wow, we get a 2% R-squared! 😉 It also shows that momentum persists, and higher average sector correlation has a similar effect to the above model, still positive on future returns.

Here is a heat map where the average 60-day past return is on the horizontal axis, and average sector correlation is on the vertical axis, and the variable displayed is frequency of occurrence.

Looks pretty average, with the two effects being fairly separate, with an odd southwestern quadrant.

Here is a heat map where the average 60-day past return is on the horizontal axis, and average sector correlation is on the vertical axis, and the variable displayed is average future 60-day returns.

Here are my tentative findings:

  • Price momentum persists.? 60 days of weakness/strength tends to beget 60 more days of weakness/strength.
  • Extremely high or low average sector correlations seem to go along with low returns.? Middling average sector correlations seem to go along with higher returns.
  • Low average sector correlations and lousy past price performance seems to beget excellent future performance.
  • High average sector correlations and lousy past price performance seems to beget lousy future performance.
  • When past price momentum is high, average sector correlation doesn’t seem to matter as much.
  • If past trends continue, average returns over the the next 60 days are around 2.5% with a very wide error band.? Current average sector correlation is 50%, and past 60 days returns are 9%.

One final graph:

Do you see a pattern here where high average sector correlations come before market peaks?? I don’t.

All that said…

I know the earlier articles dealt with asset class correlations, rather than correlations with stock market sectors.? I would have expected the same result.? Maybe there is aonther way to do this analysis separating out safe sectors from risky sectors.

But what are safe sectors?? Utilities? Consider 2001-2003.? Telephone services? Also 2000-2003.? Financials? 2007-?? Perhaps Consumer Staples is the only truly safe sector… or maybe it should be energy? Consider the early ’90s.

This is one article where I end scratching my head, but publish anyway, because:

1) My readers may help me.

2) It is valuable to know where research dead ends exist.

After all this, I don’t see average sector correlations as a valuable variable in investing.? Maybe that is not true of asset classes.? If anyone has the proper data to send to me on that, I will reproduce an analysis like this, and tell you what I find.

Theme: Overlay by Kaira