science and math vs. academic finance (pension consultants, too)

“Evaluating Trading Strategies”

The Buttonwood column in the February 21st issue of The Economist talks about a recent article published in the Journal of Portfolio Management, titled “Evaluating Trading Strategies,” authored by Profs. Campbell Harvey of Duke and Yan Liu of Texas A&M.

Long ago, I’d come to think of the difference between academic financial theorists and portfolio managers as somewhat like that between teachers of academic literary theory and actual authors.  That is to say, the two sets of people live in very different worlds, with little in common    …and without that much relevance for each other.

Three exceptions with finance:

–academics are often used as front men for various investment schemes, such as in the case of the ill-fated Long-Term Capital Management, which raised a huge amount of money to implement a strategy of buying illiquid bonds and collapsed shortly thereafter–destabilizing the world financial system in the process

–they often sit as window dressing on the boards of directors of financial companies, and

–their theories inform much of the methodology of the investment consultants on whose advice pension fund managers rely heavily.

its conclusion

The article has an emperor’s new clothes aspect to it.

Simply put, it says that academic finance researchers routinely use a standard for testing for the statistical significance of their findings that is much too weak and alrady discredited in mainstream scienticif research.  Because of this failing, in statistical work in finance large numbers of false.  This is through ignorance, not malice.

As the authors put it:

“So where does this leave us? …Most of the empirical research in finance, whether published in academic journals or put into production as an active trading strategy by an investment manager, is likely false.  …half the financial products (promising outperformance) that companies are selling to clients are false”

Who knows whether this article will have any long-term effects?

In the real world, very few people take academic finance theories seriously–except for pension funds, which rely heavily on consultants who use it to legitimize their advice.  The conclusion that the “advice” is little more than picking numbers out of a hat (arguably even less reliable than that method) has the potential to really shake up this chronically poor-performing sector.

stock buybacks: the curious case of IBM

Regular readers will know that I’m not a fan of stock buybacks by companies.  I believe that even though buybacks are advertised as returning cash to shareholders in a tax-efficient way, their main effect–even if not their purpose–is to keep the dilutive effects of management stock options away from the attention of ordinary shareholders.  Admittedly, I haven’t done a study of all firms that buy back stock, but in the cases I have looked at the shares retired this way somehow end up offsetting new shares issued to management.  As a result, you and I never see the slow but steady shift in ownership away from us and toward employees.

In recent years, activist investors have made increasing stock buybacks a staple of their toolkit for “helping” stick-in-the-mud companies improve their returns.  Certainly, accelerating buybacks can give a stock an immediate price boost.  But since I don’t believe that the usual activist suspects have your or my long-term welfare as shareholders at heart, I’ve had an eye out for cases where extensive buybacks have ceased to work their magic.

I found IBM.

Actually I should put the same ” ” around found that I put around helping two paragraphs above.  I stumbled across an article late last year in, I think, the Financial Times that asserted all IBM’s earnings per share growth over the past five years came–not from operations–but from share buybacks.  A case of what Japan in the roaring 1980s called zaitech.  Hard to believe.

I’ve finally gotten around to looking.  I searched in vain for the article.  I found a relatively weak offering from the New York Times Dealbook, whose main source appears, somewhat embarrassingly for the authors, to have been IBM market-speak in its annual report.  I did find an excellent two-part series in the FT that I’d somehow missed but which appeared earlier this month.  It’s useful not only conceptually but also for IBM history.

IBM

The FT outlines the essence of the IBM plan to grow eps from $11.52  in 2010 to $20 by this year–a target abandoned last October by the new CEO..  Of the $8.50 per share advance, $3.50 was to come from revenue growth, both organic and from acquisitions; $2.50 each were to come from operating leverage–which I take to be the effect of keeping SG&A flat while revenues expanded–and share buybacks.

What actually happened from 2010 through 2014 is far different:

–IBM’s revenues, even factoring in acquisitions, fell by 7% over the five years

–2014’s operating profit was 5% higher than 2010’s

–net profit grew by 7.0%, aided by a lower tax rate,

–nevertheless, earnings per share grew by 35%!

How did this happen?

Over the five years, until share buybacks came to a screeching halt in 4Q14, IBM spent just about $70 billion on the open market on its own stock.  That’s over 3x the company’s capital expenditures over the same period.  It’s also about 3x R&D expenditure, which is probably a better indicator for a software firm.  And it’s over 3x dividend payments.

The buying reduced the share count by 315 million to 995 million shares.  The actual number of shares bought, figuring a $175 average price, would have been about 400 million.  I presume the remainder are to offset shares issued to employees exercising stock options (although there may be some acquisition stock in there–no easy way to find that out).

results?

What I find most interesting is that, other than a flurry in the first half of 2011, the huge expenditure did no good.  IBM shares have underperformed pretty consistently, despite the massive support given by the company.  And IBM has $13 billion more in debt that it had before the heavy buybacks began.

Where is the company now?

I don’t know it well enough to say for sure, but it appears to me that it has taken recent earnings disappointments to jolt IBM into the realization that the 2010 master plan hasn’t worked.  A half-decade of the corporate equivalent of liposuction and heavy makeup has not returned the firm to health.  Instead, IBM has burned up a lot of time   …and a mountain of cash.

I think it’s also reasonable to ask how ordinary IBM shareholders have benefitted from the $60+ per share “returned” to them through buybacks.  I don’t see many plusses.  The stock dropped by about $20 last October, when IBM officially gave up the 2010 plan, so some investors were fooled by the company’s zaitech.  But spending $60+ to postpone a $20 loss that happened anyway doesn’t seem like much of a deal.

Only the board of directors knows why almost five years elapsed before anyone noticed the plan had long since gone off the rails.

Warren Buffett’s latest portfolio moves: the 4Q14 13-f

Investment managers subject to SEC regulation (meaning basically everyone other than hedge funds) must file a quarterly report with the agency detailing significant changes in their portfolios.  It’s called a 13-f.  Today Berkshire Hathaway filed its 13-f for 4Q14.  I can’t find it yet on the Edgar website, but there has been plenty of media coverage.

Mr. Buffett has built up his media and industrial holdings, as well as adding to his IBM.  The more interesting aspect of the report is that it shows him selling off major energy holdings–ExxonMobil, which he had acquired about two years ago, and ConocoPhillips, which he had been selling for some time.  Neither has worked out well.

There’s also a smaller sale of shares in oilfield services firm National Oilwell Varco and a buy of tar sands miner Suncor–both presumably moves made by one of the two prospective heirs working as portfolio managers at the firm (whose portfolios are much smaller than Buffett’s.  Buffett has told investors to figure smaller buys and sells are theirs.)

Three observations:

–the Buffett moves would have been exciting–maybe even daring–in 1980.  Today, they seem more like changing exhibits in a museum.

–if I were interested in Energy and thought it more likely that oil prices would rise than fall, I’d be selling XOM, too.  After all, it’s one of the lowest beta (that is, least sensitive to oil price changes) members of the sector.

But I’d be buying shale oil and tar sands companies that have solid operations and that have been trampled on Wall Street in the rush to the door of the past half-year or so.  That doesn’t appear to be Mr. Buffett’s strategy, however.  His idea seems to be to cut his losses and shift to areas like Consumer discretionary. (A more aggressive stance would be to increase energy holdings by buying the high beta stocks now, with the intention of paring back later by selling things like XOM as prices begin to rise.)  NOTE:  I’m not recommending that anyone actually do this stuff.  I’m just commenting on what the holdings changes imply about what Mr. Buffett’s strategy must be.

–early in my career, I interviewed for a job (which I didn’t get) with a CIO who was building a research department for a new venture.  I was a candidate because I was, at the time, an expert on natural resources.   The CIO said the thought there were three key positions any research department must fill:  technology, finance and natural resources.  All require specialized knowledge.    I’d toss healthcare into the ring, as well.  I’d also observe that stock performance in these more technical areas is influenced much less by the companies’ financial statements than is the case with standard industrial or consumer names.

Mr. Buffett is an expert on financials–he runs a gigantic insurance company, after all.  On tech and resources, not so much, in my opinion.  Financials are the second-largest sector in the S&P 500, making up 16% of the total.  Tech makes up 19.5%; Energy is 8.3%; Healthcare 14.9%.  The latter three total 42.7% of the index.  As a portfolio manager, it’s hard enough to beat the index in the first place.  Being weak in two-fifths of it makes the task even harder.

what will a soft dollar-less world look like

Yesterday I wrote about an EU regulatory movement to eliminate the use of soft dollars by investment managers–that is, paying for research-related goods and services through higher-than-normal brokerage commissions/fees.

Today, the effects of a ban…

hedge funds?

I think the most crucial issue is whether new rules will include hedge funds as well.  The WSJ says “Yes.”  Since hedge fund commissions are generally thought to make up at least half of the revenues (and a larger proportion of the profits) of brokerage trading desks, this would be devastating to the latter’s profitability.

Looking at traditional money managers,

 $10 billion under management

in yesterday’s example, I concluded that a medium-sized money manager might collect $50 million in management fees and use $2.5 million in soft dollars on research goods and services.  This is the equivalent of about $1.6 million in “hard,” or real dollars.

My guess is that such a firm would have market information and trading infrastructure and services that cost $500,000 – $750,000 a year in hard dollars to rent–all of which would now be being paid for through soft dollars.  The remaining $1 million or so would be spent on security analysis, provided either by the brokers themselves or by third-party boutiques (filled with ex brokerage house analysts laid off since the financial crisis).

That $1 million arguably substitutes for having to hire two or three in-house security analysts–and would end up being distributed as higher bonuses to the existing professional staff.

How will a firm pay the $1.6 million in expenses once soft dollars are gone?

–I think its first move will be to pare back that figure.  The infrastructure and hardware are probably must-haves.  So all the chopping will be in purchased research.  The first to go will be “just in case” or “nice to have” services.  I think the overwhelming majority of such fare is now provided by small boutiques, some of which will doubtless go out of business.

–Professional compensation will decline.  Lots of internal arguing between marketing and research as to where the cuts will be most severe.

smaller managers

There’s a considerable amount of overhead in a money management operation.  Bare bones, you must have offices, a compliance function, a trader, a manager and maybe an analyst.  At some point, the $100,000-$200,000 in yearly expenses a small firm now pays for with soft dollars represents the difference between survival and going out of business.

Maybe managers will be more likely to stick with big firms.

brokers

If history is any guide, the loss of lucrative soft dollar trades will be mostly seen more through layoffs of researchers than of traders.

publicly traded companies

Currently, most companies still embrace the now dated concept of communicating with actual and potential shareholders through brokerage and third-party boutique analysts.   As regular readers will know, I consider this system crazy, since it forces you and me to pay for information about our stocks that our company gives to (non-owner) brokers for free.

I think smart companies will come up with better strategies–and be rewarded with premium PEs.  Or it may turn out that backward-looking firms will begin to trade at discounts.

you and me

It seems to me that fewer sell-side analysts and smaller money manager investment staffs will make the stock market less efficient.  That should make it easier for you and me to find bargains.

 

 

 

the demise of soft dollars

This is the first of two posts.  Today’s lays out the issue, tomorrow’s the implications for the investment management industry.

so long, soft dollars

“Soft dollars” is the name the investment industry has given to the practice of investment managers of paying for research services from brokerage houses by allowing higher than normal commissions on trading.

Well understood by institutional, but probably not individual, clients, this practice transfers the cost of buying these services–from detailed security analysis of industries or companies to Bloomberg machines and financial newspapers–from the manager to the client.  In a sense, soft dollars are a semi-hidden charge on top of the management fee.

In the US, soft dollars are reconciled with the regulatory mandate that managers strive for “best price/best execution” in trading by citing industry practice.  This is another way of saying:   whatever Fidelity is doing–which probably means having commissions marked up on no more 15%-20% of trades.

In 2007, Fidelity decided to end the practice and began negotiating with brokers to pay a flat fee for research.  As I recall, media reports at the time said Fidelity had offered $7 million in cash to Lehman for an all-you-can-eat plan.  Brokerage houses resisted, presumably both because they made much more from Fidelity under the existing system and because trading departments were claiming credit for (and collecting bonuses based on) revenue that actually belonged to research.

theWall Street Journal

Yesterday’s Wall Street Journal reports that the EU is preparing to ban soft dollars in Europe for all investment managers, including hedge funds, starting in 2017.

not just the EU, however

Big multinational money management and brokerage firms are planning to implement the new EU rules not just in the EU, but around the world.

Why?

Other jurisdictions are likely to follow the EU’s lead.  Doing so also avoids potential accusations of illegally circumventing EU regulations by shifting trades overseas.

soft dollars in perspective

in the US

Let’s say an investment management firm has $10 billion in US equities under management.  If it charges a 50 basis point management fee, the firm collects $50 million a year.  Out of this it pays salaries of portfolio managers and analysts, as well as for research travel, marketing, offices… (Yes, 12b1 fees charged to mutual fund clients pay for some marketing expenses, but that’s another story.)

If the firm turns over 75% of its portfolio each year, it racks up $7.5 billion in buys and $7.5 billion in sells.  Plucking a figure out of the air, let’s assume that the price of the average share traded is $35.  The $15 billion in transactions amounts to about 425 million shares traded.  If we say that the manager allows the broker to add $.03 to the tab as a soft dollar payment, and does so on 20% of its transactions, the total annual soft dollars paid amount to $2.5 million.

foreign trades

Generally speaking, commissions in foreign markets are much higher than in the US, and soft dollar limitations are    …well, softer.  So the soft dollar issue is much more crucial abroad.

hedge funds

Then there are hedge funds, which are not subject to the best price/best execution regulations.  I have no practical experience here.  I do know that if I were a hedge fund manager I would care (almost) infinitely more about getting access to high quality research in a timely way (meaning ahead of most everyone else) than I would about whether I paid a trading fee of $.05, $.10 (or more) a share.

We know that hedge funds are brokers’ best customers.  Arguably, banning the use of soft dollars–enforcing the best price/best execution mandate–with hedge funds would be devastating both to them and to brokerage trading desks.

translating soft dollars to hard

When I was working, the accepted ratio was that $1.75 soft = $1.00 hard.  I presume it’s still the same.  In other words, if I wanted a broker to supply me with a Bloomberg machine that cost $40,000 a year to rent, I would have to allow it to tack on 1.75 * $40,000  =  $70,000 to (the clients’) commission tab.

 

Tomorrow, implications of eliminating soft dollars

 

 

 

 

 

 

a report card for smart beta

Purveyors of “smart beta” equity portfolio strategies have been very popular over the past few years, both with individual investors and with institutions.

The source of the attraction is clear:

smart beta claim to provide better performance than an index fund without engaging in active portfolio management. Actually, it claims to outperform because it doesn’t employ value-subtracting human portfolio managers to muck up the works.  Rather, smart beta operates by reshaping the weightings of stocks in the index according to predetermined computer-managed rules.  (I’ve written about smart beta in more detail in other posts.)

In other words, it’s free lunch.

My observation is that smart beta is a marketing gimmick  …one that has been very successful in bringing in new money, but a gimmick nonetheless.  Basically what it does is to create a portfolio that contains the index constituents, but in different proportions from their index weightings.  The rules for determining the smart beta weightings are set in advance and the portfolio is periodically rebalanced to restore the “correct” proportions.  For my money, the preceding sentence describes active management.  The portfolio managers are just hidden behind a computer curtain.

A simple example of smart beta:  maintain a portfolio of S&P 500 names but have .2% of the money in each stock–rather than having it loaded up with lots of Apple, ExxonMobil, Microsoft, Johnson&Johnson and Berkshire Hathaway.  Historically, this is a strategy that had its best run in the late 1970s – early 1980s, but which followed with a very extended period of sub-par performance.

Anyway,

I was catching up on my reading of the Financial Times over the weekend and came across an article from the FTfm of February 2nd titled,“Smart beta is no guarantee you will beat the market.”

It turns out that of the 10 biggest smart beta ETFs in the US, seven have underperformed over the past three years and five over the past five years.

That’s not that different from what active managers have done.

However, unlike the case for active managers, assets under smart beta management have grown fivefold since 2009.

 

 

 

the euro, the US$ and the Swiss franc

With the beginning of quantitative easing by the European Central Bank, the euro has slipped against the USD by about another 3% today to a value of 1 € = US$1.12.  That’s a decline in the euro of about 7.5% just since January 1st.  The EU currency has tumbled by more than 14% vs. the greenback over the past year, and by almost 20% since its high of $1.39+ last May.

This is an astounding fall for the world’s second most important currency.  It’s an enormous boost for EU-based enterprise overall and for exporters in particular–as well as a huge burden for their hard currency-based rivals. It would also be a mind-boggling loss of national wealth for EU citizens, were it not that Japan has depreciated the yen by a third over the past few years in a bid to regain global relevance for its manufacturing base.

Enough of this.   Down to brass tacks:

the euro/dollar

The income statements of US companies with EU exposure will be savaged by the currency decline.  Yes, in theory they may be able to raise prices to recover some of their depreciation-created losses.  But the general rule in this situation is that prices can only go up in line with overall inflation–which is non-existent in the EU at the moment.

My strong feeling is that Wall Street hasn’t fully worked this out yet.  So combing through our holdings to find euro victims should be a high priority for each of us.

the euro/Swiss franc (CHF)

The CHF has gained almost 25% against the euro since the Swiss central bank depegged its currency from the euro a little more than a week ago.  The speed of the move clearly shows what should have been apparent over the past year of euro depreciation–that the Swiss government was trying to maintain a peg that was miles away from where the cross rate would be without constant economy-distorting intervention.

We know this sort of thing can’t last.  If the forty-year history of floating exchange rates shows anything it’s that trying to maintain an artificial exchange rate always ends in disaster.  Yet what continues to come out in the press post-depegging is that:

–lots of EU property owners had decided it was a great idea to take out a CHF-denominated mortgage on their homes.  Short-term rates were negative, after all.  Ouch!

–a number of commodities brokers are in serious financial trouble because they allowed individual clients to build up short-CHF positions on margin that were so big there’s no chance they’ll ever be able to repay the losses they’ve incurred.

–there’s been a parade of currency trader departures from hedge funds caught out by the same short-CHF bet.

I guess this just shows that P T Barnum was right–that despite the examples of the collapse of the pre-euro Exchange Rate Mechanism in the early 1990s, the Asian debt crisis later in that decade and all of the problems with one-size-fits-all Eurobonds, there are still tons of people willing to take what history shows is the losing side of a wager.

 

 

Follow

Get every new post delivered to your Inbox.

Join 345 other followers