why September’s such a bad month for stocks

welcome to September 2011

This year, September has opened to a mini-swoon in world stock markets caused by a poor jobs report in the US, worries about the government suing banks over past sub-prime mortgage sins, and general panic about Greece (the EU political “plan,” if you’d call it that, appears to be to let the situation deteriorate to the point that voters will be grateful for even a painful rescue and not kick out the politicians who caused the problem in the first place).

the annual September equity decline

Who knows how long this downdraft will last–as I’m writing this, global equities appear to be rallying a bit, but this isn’t the normal seasonal decline in stocks.

It’s really not just September when stocks go down, either.  There’s a several-week period of selling that typically starts each year in mid-September and ends in mid-October.  But there’s usually a rally toward the end of October, so the early-month decline is less obvious.

This decline has nothing to do with the macroeconomy or stock valuation.  It’s all about mutual fund taxes.

here’s why

Mutual funds in the US (ETFs, too) are a special type of corporation.  Their activities are limited to investing, and they’re required to distribute to shareholders virtually all of their net realized profits soon after the end of each tax year.  In return for these restrictions, they’re exempt from corporate tax on their gains.  Only shareholders pay.

The tax year for virtually all mutual funds, which determines how much they must distribute, ends on October 31st.

adjusting the distribution

Shareholders like to get a distribution, which they take to be a sign that things are going well.  This makes no sense to me–better to “ride your winners” and let gains compound without paying tax–but that’s what the customers want.

On the other hand, people don’t like to pay taxes, so they don’t want a gigantic distribution (over 5% of the fund’s assets), either.

So mutual fund managers start to adjust the size of their potential distributions sometime in September.

This involves a lot of selling. 

If the required distribution is too big, a manager will scour his portfolio for stocks where he has a loss that he can sell.  If there’s no distribution, or if the payout will be too small, he hunts around for positions where he can justify taking a partial profit. 

It’s not about actually sending money to shareholders,

as I’ve heard “experts” on finance talk shows say.  An overwhelming majority of mutual fund shares, say, 95%+, sign up for automatic reinvestment of distributions.  So if the yearend gains add up to 5% of the fund assets, the amount of money that actually leaves the fund is .05 x .05 = .0025, or .25% of the assets.  That’s far less than the frictional cash a manager needs to have on hand to ensure smooth settlement of tradesSo the transfer of funds is not a big deal.

this tax planning is healthy, in my view

It gives a reason for a manager to step back and take a hard look at all the fund’s positions It also gives him a psychological excuse to dump out stocks where he’s hoping against hope that they’ll work out (trust me, even the top managers have one or two of them).

one caveat

If a fund has unused tax losses left over from prior years–and many still have them as scars from panic redemptions by shareholders in late 2008-early 2009–it can’t make a distribution until those losses are gone.  Either the fund makes offsetting gains (which won’t be subject to tax–a good thing) or the losses expire.

In either case, there’s no need to take part in the yearly September-October tax selling ritual.

this year?

My guess is that tax selling season will be relatively mildThe S&P 500 is showing about a 1% loss since last Halloween.   So unless a manager made very large adjustments to his portfolio positioning a few months ago, when stocks were considerably higher, the gains generated in day-to-day portfolio activity shouldn’t be large.  Also, at least some funds will continue to be in a net loss position, so they won’t be able to make distributions no matter what.



US corporates lobby for a new tax amnesty

Yesterday’s Financial Times contains three (count ’em, three) articles, one on the front page and prominently above the fold, talking about recent efforts by US corporations in lobbying Washington to allow them to repatriate foreign cash holdings while paying little or no income tax.

This appears to be the start of a public relations campaign by the US Chamber of Commerce aimed at persuading Congress to pass a tax amnesty bill like the Homeland Investment Act (HIA) of 2004.

US tax law, unlike that of many other countries, makes multinationals incorporated in the US pay domestic income tax (less a credit for foreign income taxes paid) on any foreign earnings repatriated here.  This can be a big deal.  For a US company recognizing Asian profits in Hong Kong, for example, the corporate tax is zero.  This means a firm bringing money like this back home would typically have to pay 35% of it to the IRS.  The funds are probably going to be reinvested in growing Asian businesses.  But even if not, unless the funds are crucially needed in the US it would be financially foolish to repatriate it.

HIA allowed firms to pay a maximum of 5.25% tax on any money brought back to the US during a specified period of time.  Companies were required to use the repatriated funds only to hire new workers or to invest in plant and equipment.  The idea was that this would reduce unemployment and spur new investment.  None could be used for stock buybacks, dividend payments or executive compensation.

According to forthcoming research, the reality of the HIA was quite different.  Corporations repatriated around $300 billion from abroad.  Strictly speaking, all the money was used for the purposes intended.  But aggregate employment and capital investment didn’t increase.  The funds simply freed domestically generated profits to be used for dividends etc..  In fact, some firms actually used the repatriated funds to replace domestic profits that they shipped abroad.

Proponents of  HIA II, which the FT says include Cisco, GE and Microsoft, are not making strong claims this time around.  They label the cash, which is estimated at about $1 trillion, as being “trapped” abroad.  They argue that maybe $400 million would be repatriated under HIA II, giving the government $20 billion or so in tax money it wouldn’t otherwise have.  And the repatriated funds would likely slosh around doing something–presumably economically good–in the US.

So far the Obama administration is saying no, seeing that it’s in enough trouble without advocating a big tax break for cash-rich corporations.

investment implications (there actually are some)

1.  I wrote about this topic a bit last April, specifically regarding the large buildup of cash on the balance sheets of technology companies.

2.  To be able to pay it out in dividends, a US-incorporated company has to have the cash available in the US.  But most publicly-traded companies don’t disclose enough about where there cash balances are, or the cash generating/cash using characteristics of their US and foreign businesses for an analyst to see how well a given dividend is covered.  This didn’t make any difference when dividend yields were very low and investors were interested in capital gains.  But it does now.

3.  It takes a US$1.50 earned pretax in the US to give the same lift to the reported earnings of a US company as US$1 earned in Hong Kong.  So a corporation concerned with maximizing eps might well choose to recognize profits in Hong Kong rather than the US, assuming it had a choice.  Similarly, a decision to shift the profit stream to the US in order to may dividends–again, assuming this were possible–would mean a structurally lower level of eps.  I suspect that at some point, investors will ask for earnings estimates that are “normalized,” in the sense of adjusted to what they would be under a standard 35% tax rate, in order to get a more apples-to-apples comparison.

4.  Why lobby for HIA II?  The paper I linked to above argues that it makes very little difference to corporations in the aggregate.  But there may be firms–most likely in the tech area–who will be forced to borrow or to repatriate foreign cash balances (and pay tax on them), either to be able to maintain the current dividend or raise it.  The strongest advocates of a new HIA might well be in this position.

The Lehman Report, “Repo 105,” and “tobashi”

The Valukas report

A court-appointed examiner, Anton Valukas, released his nine-volume, 2200 page report on the bankruptcy of Lehman Brothers last Thursday, after more than a year of investigation.

I haven’t read the report and I don’t intend to, since I’m pretty sure it won’t be chock full of useful investment information.  There’s one aspect of the newspaper accounts of Valukas’s work that jumps out to me, though–the now-becoming-infamous “repo 105″ transactions.  It isn’t just that Lehman actively distorted its financial statements so that Wall Street would not understand the true extent of its borrowings.  It’s that all the distortions emanated from London.

Why do US transactions in London?

From the beginning of the financial crisis it has struck me as odd that very many of the “toxic” asset transactions done by the big commercial and investment banks were executed in London–even though they involved US assets, US-based sellers and US-based buyers!

nothing by accident

It’s my experience that nothing big companies do happens by accident.  There’s always a reason, even if you can’t immediately see what it is.  In the toxic asset case, I thought the two logical possibilities were that:

–there was an economic reason–a tax advantage, perhaps, or lower execution costs–to doing the transactions in the UK, or

–there was legal one–firms were trying to protect themselves from civil or criminal action.  In other words, they were doing things that London’s “regulation lite” philosophy might lead it to turn a blind eye to, but which would be clearly illegal in the US.

The Lehman Report seems to tip the balance in favor of the second explanation.

What “Repo 105” was

The name has already caught reporters’ fancy.  In its simplest form, toward the end of each quarterly reporting period Lehman would agree with commercial banks:

(1) to exchange large baskets of the company’s assets for cash, and

(2) to repurchase the assets at a higher price a few days later, after the end of the quarter.

Lehman would use the proceeds of these “repurchase agreements” to reduce the debt outstanding on its balance sheet at quarter’s end.  Repos are very common, plain-vanilla transactions in finance.  A money market fund, for example, might buy a short-term note from, say, IBM for 99 that both sides agree will be cashed in a month from now for 100.  So the fact of repos or even that they made debt “vanish” for a few days is not where the problem lies.

But Lehman also decided that in its SEC filings and other official reporting to shareholders, it would suppress the information about its repurchase obligations.  By only showing one side of the trade, it made itself seem to have less debt than it actually had.  In its last year of existence, Lehman was hiding close to $50 billion in borrowings.

Illegal in the US, ok in the UK

Lehman maintained that if it was fancy enough in structuring the transactions, it would get away with not disclosing the repurchase obligations.  Ernst & Young, Lehman’s auditors, had no quarrel with this.  But Lehman couldn’t find a single US law firm willing to say that doing so was legal.  Put another way, every law firm it approached told Lehman what it was proposing to do was against the law.

How did Lehman respond?

It found a British law firm willing to say that not revealing the repurchases would be legal in the UK–and then did all the transactions in London!

As Lehman got into deeper and deeper trouble, the amounts repoed got larger.  During 2008 the repos approached $50 billion, enough to “lower” Lehman’s financial leverage (its borrowings divided by net worth) by over 10%.  That’s an enormous difference.

Madoff redux

Even though the amounts were mammoth and that reducing leverage was one of his key aims, Lehman’s chairman, Richard Fuld, reportedly denies any knowledge of the scheme.  And in a reprise of the Bernie Madoff scandal, a very persistent whistleblower was apparently ignored by regulators, and Lehman’s top management and board of directors alike.

It will be interesting to see if the Valukas report is an effective counter to the intensive, and so far successful, lobbying efforts of the banks to maintain the status quo, avoid prosecution of their managements, and stymie regulatory reform.

We’ve seen this once before–“tobashi”

During the second half of the Eighties, when the Japanese stock market was booming, investment bankers persuaded many domestic companies to raise capital by issuing bonds with warrants attached.

The companies didn’t really need the money, but the bankers’ sales pitch was persuasive:  give the money to us, they said, to invest for you in the Japanese stock market.  As the Nikkei rises, we’ll make lots of money for you.  And you’ll pay the bonds back with the funds you’ll get when the warrant holders exercise their rights to buy new shares of your stock at much higher prices than today’s.  You win two ways.

Events didn’t work out as planned, though.  For one thing, most of the warrants expired worthless, leaving the issuing companies stuck with repaying bondholders.

Just as bad, the Japanese investment banks lived up to their reputations as notoriously bad investors by losing in the stock market virtually all the money entrusted to them in the corporate stock accounts.  That’s when they came up with the idea of “tobashi,” or “hot potato,” as a way of disguising from company shareholders the fact of their horrible investment performance.

Let’s say the original amount invested, and the carrying value of the portfolio, was 100, but the money left was only 10.  (Hard as it is to imagine, I think this would have been a typical situation.)  The investment banks would find a third party willing to “buy” the portfolio at a price of 100 just before balance sheet date and sell it back to the original holder for the same amount a couple of days later.  Thereby, the losses would never be discovered.  This activity was a very closely guarded secret.

What eventually happened?  One reporting period, a company decided that selling a portfolio worth 10 for 100 was a great deal.  So it refused to accept back the “hot potato” it had tossed to the other firm.  The whole fraudulent scheme quickly became public.

Maybe this is where Lehman got the idea.  After all, it was around in Japan at the time.

Mutual fund cash levels: what they mean

The Investment Company Institute, the trade association for mutual fund management companies, just issued a periodic report on mutual fund inflow and outflows.  Among other things, it shows that the percentage of cash held by equity fund managers–with both international and US-only mandates–dropped over the past year from 5.7% to 3.6%.  This is a shrinkage in dollar terms from $210 billion to $173 billion, despite a rise in overall assets during the twelve months.

Bloomberg, citing Wall Street strategists, says this is a bad thing.  Why?  In their view, it’s because a low cash percentage signals an impending market decline.  Noting this, investors stop aggressively buying stocks.  Huh?

I don’t think this is right, for several reasons. Continue reading