Asymmetrical Information

I don't see that Boksin's made any error as far as his discount rate is concerned. That seems standard practice.

I'm more concerned about his assuming a 27.7% average tax rate on distributions from retirement plans (not 28.7% as Sloan says).

IIRC the average (not top marginal) federal personal income tax rate today is closer to 15%. These retirement accounts will generally be cashed in after retirement, when there is no salary income, and such distributions get taxed starting at 0% then up through the brackets. So 27.7% as an average rate seems to me like it could be seriously too high.

But I haven't read Boskin's whole paper and how he figured that rate, so I won't lodge a formal complaint yet.

"I don't see that Boksin's made any error as far as his discount rate is concerned. That seems standard practice."

Look again, Jim.

Suppose I invest $1000 in some set of risky securities, with an expected return of 7.5%. Then I discount the returns at 5%. The PV of my $1000 jumps miraculously, at least on the spreadsheet. So I should be able to turn right around and sell my investment for more than $1000. Right?

That's nonsense, of course. Financial market investments are zero NPV investments by definition.

"That's nonsense, of course. Financial market investments are zero NPV investments by definition."

A stock portfolio has zero net present value? I think you mean that in efficient markets it has the same present value as a Treasury portfolio (ie zero excess).

Mindles,

I said NPV, not PV. All that means is that the appropriate discount rate is exactly equal to the expected return. Just like a treasury portfolio, which also has zero NPV.

Correct. Retracted. D'oh! Although efficient markets still have to be assumed for the NPV to be zero.

Discounting a securities portfolio is a very strange idea, isn't it? Market value being a discounted value & all.

I'm not sure what you mean when you say "discounting a securities portfolio is a strange idea."

Market value is just the PV of future cash flows, whether we're talking about a single security or a portfolio. Maybe I'm missing your point.

Nothing odd, I'm simply saying that as a money manager for 15 years I've never seen anyone go through the circular process of discounting the future value, at an assumed return, of a portfolio of traded securities. (corporate cash flows?, bond flows? yes, of course, but not this).

I'll keep pointing out, though, that market price is only the PV of future cash flows (hence the correct market value) if the market is efficient. Big if, for some folks.

"Look again, Jim."

Well, I've looked part-way and as far as GM goes I don't see what's got Sloan so hyped about its running a "delusional financial magic perpetual profit machine". It looks like rather conventional arbitrage to me.

As he himself points out GM gets a 2.5 point swing by paying deductible interest and collecting tax-free pre-tax interest -- which is real enough, no delusion there -- and it is going to be compounding its investment returns while paying simple interest.

Moreover it is borrowing cheap, when rates are at a 40-year low. *Presumably* rates will be higher on average over the next 30 years, adding to its arbitrage spread.

"Can GM guarantee it will earn 9 or 8 percent on the assets in its pension fund? Obviously not."

Of course it can't. And if average normal market returns over the next 30 years taken tax-free and compounded don't exceed today's historic-low borrowing cost even after interest is deducted for taxes, GM surely will be in serious trouble. But, hey, let's face it -- in that case we'll *all* be in serious trouble.

So how he figures that in GM's case this is a "perpetual-motion machine" that can be expected to produce "paper profits but no real money" I don't know.

As to Boskin's case I will defer comment until I read his whole paper.

The ratio 7.5 to 5.3 is sufficiently close to 70%, and 70% is sufficiently close to 1 minus the marginal rate of tax on investments to make me think that what we are actually seeing here is the fact that the government doesn't pay tax; Boskin is assuming that the stock market has a post-tax rate of return, but is using a pre-tax discount rate to discount the cashflows. Which strikes me as potentially the right thing to do.

Boskin's discussion of the discount rate starts on page 88 of the working paper linked on this site earlier; I haven't time to read it for conclusions, but it looks at first glance as if he's certainly considered these issues.

I'm not familiar with pension accounting and taxes, so it's hard for me to see exactly what's happening here, but it does seem odd.

To simplify, suppose GM had just issued the 7.5% bonds directly to the pension fund. Then are we saying that GM would pay the 7.5%, record it as a deductible expense, and then turn around and record it as non-taxable income, making a magical 2.5%?

If so, why doesn't everybody do this?

And by the way, why is the coupon rate on the bonds the correct number to use in calculating GM's cost? And doesn't all this tell us that GM is a lousy investment, since they have no better use of cash?

Did I just answer my own question?

"I'm not familiar with pension accounting and taxes ... To simplify, suppose GM had just issued the 7.5% bonds directly to the pension fund..."

Can't do that. Illegal. If GM then defaulted on its bonds its retirees would be screwed. Your own IOU is not real savings and is not real security for a debt you owe. (see "Social Security Trust Fund".)

GM must deposit cash out-of-pocket in the pension fund, which the fund will invest. GM can borrow to get the cash, but then if it defaults on its debt GM's shareholders and current workers get screwed, rather than its retirees.

"Then are we saying that GM would pay the 7.5%, record it as a deductible expense, and then turn around and record it as non-taxable income, making a magical 2.5%?"

It works like this: GM borrows at 7.5%. The interest is deductible so the after-tax cost is 5% (taking Sloan's numbers). GM then contributes the borrowed funds to the pension plan trust. The trust invests them in whatever to get, let's say for illustrative sake the exact same 7.5%. This income is tax-free to the trust. (And it is also beyond GM's reach. GM is not paying this money to itself.)

Now, GM is required by law to make contributions to the pension plan each year to keep the trust funded at a certain actuarial level. These contributions reduce GM's earnings just like any other operating expense.

By borrowing to finance a contribution GM gets the benefit of tax arbitrage. For each dollar borrowed, next year (a) it will pay 5 cents interest cost after tax, but (b) the pension plan's assets will go up by 7.5 cents, reducing GM's need to make a contribution to the plan by that much -- thus saving it 2.5 cents of expense, and increasing its bottom-line earnings by that much. Moreover, the 5% is simple interest but the 7.5% compounds, so the boost to GM's bottom line increases each year.

Of course in reality GM is intending real arbitrage as well as tax arbitrage. It expects the pension trust to earn a good deal more than 7.5% because (1) it will be investing for the long run in a mix of assets that should earn more than just bonds, and (2) GM expects average rates will be higher over the next 30 years than today's 40-year low rates, even on bonds. The higher the rate of return the pension trust earns in the future, the bigger the compounding benefit to GM.

"If so, why doesn't everybody do this?"

Everybody does do tax arbitrage, it's routine. People do it with their own IRAs and 401(k)s.

Companies don't often borrow $10 billion to make a pension plan contribution because it is a massive expansion of the liability side of the balance sheet -- that's real debt there -- and leverage is risky.

But not many companies are as far in arrears on their pension obligations as GM is -- it's also doing this to make its unions and the pension regulators happier. GM's finance people are betting that with rates at a 40-year low they have a chance now to "borrow cheap" to leverage themselves out of a deep hole while making the people they have to deal with happy.

However, there's nothing at all "magical" or "perpetual motion machine-ish" about it. It's just a combination of arbitrage and leverage. You can see it on Wall Street a hundred times a day. So I think Sloan's critique is a clear "miss" in trying to paint it as something other. And having now read Boskin, I think Sloan misses by as much there too.

"... doesn't all this tell us that GM is a lousy investment, since they have no better use of cash?"

?? They're very short cash. That's their problem -- why they are borrowing so much of it now. They've been unable to meet their pension funding costs in recent years and are afraid they won't be able to in the future either, from earnings, so they are taking this big step now. I mean, they are getting seriously afraid.

Lots of people think GM and Ford are indeed seriously troubled firms, in no small part because they stand to be overwhelmed by future retiree costs that their operating businesses won't be able to support. Last I looked Ford's bonds were near junk level.

But whether that makes their stocks a lousy investment depends on how they're priced, of course.

Ah, D^2 and I agree on something, for a pleasant change.

Yes, Boskin gives the whole "discount rate" issue ample consideration, giving other plausible discount rates and their effects for those who might prefer them. And he does treat these retirement accounts as if the government is a partner in owning them.

But his "simple calculation" of an average tax rate of 27.7% still is a puzzle to me.

The actual real-life average tax rate as most recently reported is 16.1%. And Boskin himself says in the paper that retirees have a lower average tax rate than the overall average rate, since the overall average is pushed up by people in their highest-earning working years.

I mean, today's 27.7% average income tax rate is at $250,000 of taxable income (*after* all deductions, etc., are taken against income) on a joint return and at about $170,000 on a single return. And tax brackets are indexed for inflation.

So either he's saying that all those future retirees he's talking about will on average be very, very rich (so we really will have no problem with Social Security and Medicare) or I'm missing something obvious here. Or he is.

"I'm not familiar with pension accounting and taxes ... To simplify, suppose GM had just issued the 7.5% bonds directly to the pension fund.."

Can't do that. Illegal. If GM then defaulted on its bonds its retirees would be screwed. Your own IOU is not real savings and is not real security for a debt you owe. (see "Social Security Trust Fund".)

Yes. I know that. I was trying to simplify. But it is logically equivalent to using the proceeds to buy identical bonds from another company.

Moreover, the 5% is simple interest but the 7.5% compounds, so the boost to GM's bottom line increases each year

The 7.5% compounds only if it is retained and reinvested at 7.5%. But if it is paid out as benefits, replacing benefit payments from company earnings, it does not compound.

And the “tax arbitrage” only works if there is compounding. Suppose GM has a $75 pension payment to make a year from now. The payment is tax-deductible, so the after tax cost is $50, just like with the scheme you describe. Aren’t they just substituting a deductible interest payment for a deductible pension payment?

And assigning a 7.5% cost to the borrowing is not quite accurate. Why doesn’t it drive up GM’s marginal cost of capital?

“Of course in reality GM is intending real arbitrage as well as tax arbitrage. It expects the pension trust to earn a good deal more than 7.5% because (1) it will be investing for the long run in a mix of assets that should earn more than just bonds, and (2) GM expects average rates will be higher over the next 30 years than today's 40-year low rates, even on bonds. The higher the rate of return the pension trust earns in the future, the bigger the compounding benefit to GM.

This is not arbitrage at all. It is simply making risky investments with borrowed funds. It’s no different than buying stocks on margin.

Everybody does do tax arbitrage, it's routine. People do it with their own IRAs and 401(k)s.

I would call this minimzing taxes on savings, not arbitrage.

Companies don't often borrow $10 billion to make a pension plan contribution because it is a massive expansion of the liability side of the balance sheet -- that's real debt there -- and leverage is risky.

But the pension plan underfunding is itself a liability. They are just substituting one for the other. The correct technical term, I believe, is “borrowing from Peter to pay Paul.”

In fact, what GM has done is this. They’ve borrowed a massive amount of money and invested it in the securities markets. That’s all. They are gambling that the markets will yield a sufficiently high return to make them a winner. This is leverage, not arbitrage, and as you say, it’s risky.

This suggests GM is a lousy investment precisely because the best use they have of cash is securities market investments. They don’t have positive NPV internal investments. So why buy stock in the company? At best it’s just a high-overhead index fund.

"The 7.5% compounds only if it is retained and reinvested at 7.5%. But if it is paid out as benefits, replacing benefit payments from company earnings, it does not compound...."

Self evidently. GM is investing for the long run here, obviously, not to fund next year's pension payout.

"'Everybody does do tax arbitrage, it's routine. People do it with their own IRAs and 401(k)s.'
"I would call this minimzing taxes on savings, not arbitrage."

Many people borrow to make contributions to these plans and profit from the tax swing. Tax arbitrage.

"But the pension plan underfunding is itself a liability. They are just substituting one for the other. The correct technical term, I believe, is “borrowing from Peter to pay Paul.”"

It is rather more like "refinancing a liability", which is very much "borrowing from Peter to pay Paul" of course -- but hardly necessarily unwise or non-beneficial for being so.

There's no way to pay down an existing liability without incurring the offsetting cost of either reducing assets or increasing another liability correspondingly -- so they've got to substitute for each other, Peter and Paul, since each has the same net effect on the balance sheet which must continue to balance. So one must choose between them. Thus, the issue is to make the best choice between Peter and Paul under the circumstances. (And as I mentioned before, there are also operational business [non-financial] considerations in making the choice.)

"They’ve borrowed a massive amount of money and invested it in the securities markets .... gambling that the markets will yield a sufficiently high return to make them a winner. This is leverage, not arbitrage ... it’s risky."

As I said, it's both arbitrage *and* leverage -- the 2.5 points is tax arbitrage pure and simple, and over 30 years that compounds to something.

As to being "risky", well, every course of action is risky, so one must find the course of least risk. What is the risk that...

(a) the pension trust fund will earn compound market returns over the next 30 years that when taken tax-free fail to cover the after-tax 5% simple interest cost incurred at today's 40-year low rate? Versus the risk that ...

(b) If GM hadn't borrowed as it has, in the future it would be forced to fund the same huge pension liabilities either by borrowing at a higher rate than today's historic low rate, or by taking operating funds from the business at an opportunity cost higher than today's historic-low 5% borrowing rate? (And that its shareholders will then look back and want to put the head of today's CEO's on a pike?)

They are both risks, and one must choose between them if in GM's place. GM has made its choice, which would you pick?

"This suggests GM is a lousy investment precisely because the best use they have of cash is securities market investments."

??? You keep saying this. GM *lacks* cash, which is why it is borrowing so much to refinance its huge liabilities under compulsion of economic necessity.

Now, *that* may indicate it is a lousy investment of course. But it's not like GM's got a lot of excess cash left over after paying all its bills that it is investing in securities because it can't think of any way to use it in the business.

That's actually a much better description of Microsoft from the day it went public -- though in retrospect its business didn't do too badly in spite of that warning sign.

Jim,

GM is meeting its pension liabilities out of current earnings and asset sales. The earnings on this money will replace that. There is no "tax arbitrage" going on. Forget it.

Self evidently. GM is investing for the long run here, obviously, not to fund next year's pension payout.

What you consider "self-evident" is just factually incorrect.

http://biz.yahoo.com/rf/030627/autos_gm_pension_1.html

"The automaker has been selling off non-core assets and diverting cash and stock in its Hughes Electronics Corp unit to its pension and retiree health care plan."

In other words, they have huge expenses related to retirees. These expenses are every bit as deductible as interest costs. To the extent the return on the new money put in the pension plan substitutes for these expenses, which is the whole idea, the tax position is exactly the same. They pay interest instead of benefits.

Many people borrow to make contributions to these plans and profit from the tax swing. Tax arbitrage.

Only if the interest on the loan is tax-deductible, and even then it's no different than moving savings from a taxable account to a retirement account. Tax minimization.

There's no way to pay down an existing liability without incurring the offsetting cost of either reducing assets or increasing another liability correspondingly -- so they've got to substitute for each other, Peter and Paul, since each has the same net effect on the balance sheet which must continue to balance

Yes. I understand. But then why did you refer to the bond sale as "a massive expansion of the liability side of the balance sheet?"

??? You keep saying this. GM *lacks* cash, which is why it is borrowing so much to refinance its huge liabilities under compulsion of economic necessity.

Nonsense. They don't lack cash at all. They just went out and got many billions of dollars. What I'm saying is that if they had good internal opportunities they would use the borrowing there, rather than putting it in the securities markets.

As to being "risky", well, every course of action is risky, so one must find the course of least risk. What is the risk that...

(a) the pension trust fund will earn compound market returns over the next 30 years that when taken tax-free fail to cover the after-tax 5% simple interest cost incurred at today's 40-year low rate? Versus the risk that ...

No tax arbitrage. No compounding. See above.

(b) If GM hadn't borrowed as it has, in the future it would be forced to fund the same huge pension liabilities either by borrowing at a higher rate than today's historic low rate, or by taking operating funds from the business at an opportunity cost higher than today's historic-low 5% borrowing rate? (And that its shareholders will then look back and want to put the head of today's CEO's on a pike?)

You're the one talking about risk of leverage. GM's strategy may be perfectly sensible, given the hole they are in. That's not my point. My point is that they are gambling, (not totally with their own money), to get out of that hole. They are investing borrowed money in the financial markets. It might well work, but that's really all they are doing.

As for Microsoft, it has no significant debt, and maintains large cash balances. According to the 10-Q, this practice,

"reflects Microsoft’s views on potential future capital requirements relating to research and development, creation and expansion of sales distribution channels, investments and acquisitions, share dilution management, legal risks, and challenges to Microsoft’s business model."

In other words they have uses for the money. They hardly rely on gains from investing in equity markets to meet their expenses, as GM is doing.

"GM is meeting its pension liabilities out of current earnings and asset sales. The earnings on this money will replace that. There is no "tax arbitrage" going on. Forget it."

GM has *not* been meeting its pension plan liabilities -- which is why it was $19.3 billion behind on them at the end of last year.

The bond sale it is going to finance maybe a $10 billion contribution to the pension trust to make up part of that. And, as even Sloan saw, when you finance a $10 billion payment by borrowing that amount at X% interest and effectively getting a return of X+2.5% -- the 2.5 point difference being due to interest paid being tax deductible, and interest received being tax free -- that's tax arbitrage.

"'GM is investing for the long run here, obviously, not to fund next year's pension payout.'
"What you consider "self-evident" is just factually incorrect."

Aw, c'mon. You don't really think GM is borrowing $10 billion "to fund next year's pension payout" do you?

The $19.3 billion shortfall is actuarially determined over *the lifetimes of the entire workforce* -- i.e. several decades. It is a *long term* issue, like 50 years. The fund has $73 billion in assets right now. It is not PAYGO like Social Security.

The fund's investment portfolio lost $10 billion last year due to the stock market fall, and this is what the bond sale is meant to make up. These are *long term investments*.

"These expenses are every bit as deductible as interest costs."

When GM borrows funds to finance a contribution to the pension plan it gets the deduction for interest AND a deduction for the contribution to the plan. Two deductions.

"To the extent the return on the new money put in the pension plan substitutes for these expenses, which is the whole idea, the tax position is exactly the same. They pay interest instead of benefits."

Not at all. Two deductions > one deduction.

More to the point, GM is effectively just refinancing a debt to lower the interest rate on it. Like home owners are doing all over. This is no more radical than that, so I don't see what all the hullabaloo is about.

First, remember GM has *nothing* to do with making actual payments of pension benefits to its retirees -- that is done by the pension trust, which is a separate legal entity with its own management, and $73 billion of its own funds at the moment.

But GM owes payments to the trust to fund *future* retiree benefits actuarially, over the next 50 years or so. GM isn't behind on *next year's* benefits -- it's behind on a 50+ year actuarial balance.

So GM owes a deficiency to the trust fund, and this deficiency effectively *carries interest*. Say the trust fund actuaries say the fund needs to receive its annual contributions from GM *and* earn a return of 8% on them compounding in coming years to be actuarially sound (which is what the trust invests to attain). Then that's the interest rate paid by GM on deficiencies -- because GM must make up the deficient principal payment *and* the compound interest not earned on it, paying them out of future earnings, to make the trust fund whole.

But GM today can borrow at a mere 5% after-tax to plug some of this deficiency. So it has decided to refinance the pension debt that it expects to cost a future compounding 8% by paying it down using funds from a new borrowing that costs it only a simple 5%.

And that's all there is to it. What's so controversial about that? It's basically like someone paying down an 8% consumer loan with a 5% home equity loan. It boosts the bottom line. Even Sloan gives a numeric example of how it will do so if the trust fund earns 9% -- he just doubts it will get the 9%.

"'??? You keep saying this. GM *lacks* cash, which is why it is borrowing so much to refinance its huge liabilities under compulsion of economic necessity.'

"Nonsense. They don't lack cash at all. They just went out and got many billions of dollars. "

You mean: they don't lack credit (yet) so they can borrow cash they need but don't have. They've certainly lacked free cash from earnings needed to avoid having to borrow to meet this basic expense.

"What I'm saying is that if they had good internal opportunities they would use the borrowing there, rather than putting it in the securities markets."

They are NOT putting the borrowed money in the securities markets, they are paying down a major debt with it. Any securities that the trust buys with the money will be the trust's property, not GMs. They are separate legal entities. GM's unions will assure you of that.

"'As to being "risky", well, every course of action is risky, so one must find the course of least risk. What is the risk that...

"'(a) the pension trust fund will earn compound market returns over the next 30 years that when taken tax-free fail to cover the after-tax 5% simple interest cost incurred at today's 40-year low rate? Versus the risk that ...'

" No tax arbitrage. No compounding. See above."

Yes tax arbitrage, yes compounding. See above. See Sloan's example.

"'(b)If GM hadn't borrowed as it has, in the future it would be forced to fund the same huge pension liabilities either by borrowing at a higher rate than today's historic low rate, or by taking operating funds from the business at an opportunity cost higher than today's historic-low 5% borrowing rate? (And that its shareholders will then look back and want to put the head of today's CEO's on a pike?)'

"GM's strategy may be perfectly sensible, given the hole they are in. That's not my point."

Well, if you think GM's action is perfectly sensible, then what *is* your point in all this?

"My point is that they are gambling, (not totally with their own money), to get out of that hole. They are investing borrowed money in the financial markets. It might well work, but that's really all they are doing."

Get this clear: GM is NOT investing the borrowed money in the financial markets. It is paying down its debt to the pension fund with that money.

The pension trust is going to invest its money, $73 billion or so, in the financial markets whatever GM does.

GM's entire "gamble" is its belief that the pension trust will earn >5% on its investments in future years -- in which case GM benefits by borrowing at 5% to make a big contribution to the trust now. Else GM doesn't benefit.

Saying this is a "gamble" is no point at all, because it's also a gample that the pesnion trust won't earn more than >5%.

So there is no "no gamble" alternative -- unless you can name one?

If you can't then you still have to choose between (a) and (b) if you want to make a meaningful point. Which is the *best* gamble?

As for Microsoft, it has no significant debt, and maintains large cash balances. According to the 10-Q, this practice,

"reflects Microsoft’s views on potential future capital requirements relating to research and development, creation and expansion of sales distribution channels, investments and acquisitions, share dilution management, legal risks, and challenges to Microsoft’s business model."

In other words they have uses for the money. They hardly rely on gains from investing in equity markets to meet their expenses, as GM is doing.

~~~~~
Gee. You criticize GM for investing in securities markets saying....

"This suggests GM is a lousy investment precisely because the best use they have of cash is securities market investments."

... even though GM isn't investing the cash at issue in securities markets, but is *actually* using it in the business to pay down a vital business debt to its pension trust.

Then regarding Microsoft, and its putting ever more cash in the securities markets since the day it went public, you say that's good because its 10k boilerplate says it might someday find some reasons to use its cash -- even though its *actual* best use of a mere $46 billion of cash accumulated over 20 years is in securities market investments, which is "precisely" what suggests GM is a lousy investment.

I can't figure it. This is why I'll never be a successful investor. ;-)

There seems to be a curious strain of thought in Sloan's article and some of the comments on it that a firm (or individual) can't meaningfully improve its long-run financial status by exchanging a low-yield security for a higher-yield one when both have the same market value.

After all, if both securities have a market value of $X then nothing can be said to be pre-funded by the swap. The difference in yields just indicates the higher *risk* of one and the higher discount rate applied to its expected future returns that reduces its value to equal that of the other. Thus, to claim any real gain results from the exchange is the equivalent of claiming "magic" at work, or perpetual-motion-machinism.

But this is an error, as can be clearly seen if one thinks of "cash". After all, currency is a security with a fixed yield that happens to be 0%. If one can't improve one's long-term financial position by exchanging two assets of equal market value to take advantage of the higher yield of one, then one can't improve one's expected financial position 30 years from now by exchanging currency with current market value of $X for a portfolio of stocks and bonds with current market value of $X. Thus, pension plans and the rest of us too should be as happy to finance our needs of 30 years from now by holding currency as by investing in anything else. Which obviously is not true.

The thing is, there are different kinds of risk that matter with different importance in different situations. E.g., there is volatility risk (the amount that an investment's value may vary in the short run) and what may be called target appreciation risk (the risk that an investment's value will fail to grow to a specific dollar amount in the long run).

Volatility risk is greatest with speculative stocks and diminishes through short-term bonds to reach zero with currency. Target appreciation risk is the reverse -- diversified stocks are the least risky because their higher average return when compounded over long periods dwarfs short-term volatility risk in dollar terms, so they are most likely to appreciate to any given dollar amount. Bonds are less likely to do so because of their lower return, in spite of their lower volatility. And cash has zero chance of appreciating in dollar terms at all.

The discount rate (yield) that gives a security its market value reflects all kinds of risks as considered by all investors. But not all kinds of risk are relevant to specific kinds of investors. When investing for 40-years from now volatility risk is irrelevant and target appreciation risk matters, so cash and low-yield gov't securities are less attractive than diversified higher-yield investments. But for someone investing a specific amount of cash that must absolutely be available next year the reverse is true.

IOW, different risk measures for different folks. So it is a mistake to say a party can't improve its long-term situation by going short on 5% bonds and long on 8% investments of equal market value -- that it's just a "gamble" -- just as it is to say one can't improve one's long-term situation by trading cash for positive yield securities of equal market value, that that's just a "gamble".

Collapsing all risk measures as considered by everybody into the one used to determine the current market value of an investment is correct for one purpose -- determining the current market value of an investment. But there are other purposes for measuring risk out there, and other measures of risk for them.

Jim,

As I understand it, what Sloan is objecting to is the notion that GM can just assume a 9% rate of return and claim a "profit" based on that assumption.

Similarly, GM may install new manufacturing equipment expecting it to increase profits. But it shouldn't book those until it happens.

It is quite possible to increase the PV of your portfolio by changing your investments if you are not on the efficient frontier. To answer one question, note that an all-cash portfolio is not efficient, so exchanging cash for a portfolio containing a diversified set of securities, and possibly some risk-free securities, does increase PV.

It is when you already have an efficient portfolio that you increase expected return only at the cost of greater risk, and this leads to the situation you describe, where the NPV must be zero.

This does NOT mean that increasing risk is a foolish strategy. It might work out fine, and lead to improved results. But you can't just book those results today.

And it is a "gamble" in the sense that the financial outcome is uncertain. Good bets are gambles too.

Jim,

Your "volatility risk" and "target appreciation risk" are the same thing. An investment whose return has a one-year standard deviation of 20%, say, has a twenty-five year std deviation of 100%.

"As I understand it, what Sloan is objecting to is the notion that GM can just assume a 9% rate of return and claim a "profit" based on that assumption."

Well, reducing a cost adds to profits -- or reduces losses -- doesn't it?

GM owes a debt to to the pension trust. The effective interest rate on the debt is the pre-tax rate of return earned by the trust on its investments, because that's what GM will have to pay out of pocket to make the trust whole.

If GM pays off that debt by financing it at 5% interest, and the pension trust earns more than 5%, GM comes out ahead. It improves its bottom line by reducing a cost -- by paying only 5% to bondholders rather than paying 8%, 9%, whatever, to the pension trust -- increasing its net profit or reducing its net loss.

And GM doesn't book that result until it actually happens, year by year. But in its business planning it can project the difference between the 5% and the expected return in the trust, figure the projected difference on its bottom line and act accordingly.

Sloan for some reason considered parlaying two different interst rates this way to be some sort of risky magic perpetual motion machinism. But it's just refinancing a debt.

Of course everything in finance carries risk. To someone seeking income buying AAA corporate bonds rather than holding cash is a is risk because the next Great Deflation could soon arrive to provide a high return to cash and wipe out the bonds. It's happened before. But how likely is it?

"Gamble" is a word that colloqually implies unnecessary risk. As such, it doesn't seem to fit the risk-minimizing course of action (e.g., averting the risk that to pay off the same debt in the future GM would be forced to borrow at more than 5% or take funds from the business at higher cost). But if you wish ... different words for different folks.

BTW, GM took advantage of the opportunity to borrow cheap by upping the bond issue to $17 billion and says it is going to be investing several billion of the proceeds in its operating business. Microsoft is talking about maybe paying a $10 billion special dividend.

Jim,

I believe GM effectively books the profit based on the assumed return. It does not wait until year-end to see what the actual return (and consequent reduction in its required payments to retirees from operations) was.

I agree it's not very likely that your AAA bonds will be wiped out. So does everyone else.That's why GE pays only a slightly higher rate than the Treasury. Joe's Bar and Grill, in turn pays more than GE because of the greater risk.

The point is that, from the investors' side, excess expected returns are bought at the price of assuming more risk.

"volatility risk" and "target appreciation risk" are the same thing. An investment whose return has a one-year standard deviation of 20%, say, has a twenty-five year std deviation of 100%.
~~~

You are talking of just one kind of investment. But if I have $100x to invest and a target of $150x inflation-adjusted in ten years, what is the risk that I will fail to achieve this target if I invest the $100x in diversified stocks, government bonds, or cash respectively? The lower volatility risk of cash does not appear to reduce the risk of missing the target.

The three investments have the same current market value in exchange, but they do not have the same value for the purpose of meeting the future target. That's all I'm saying -- it's not a very profound observation.

And I note it only in context of what seems to me to be Sloan's curious objection to the idea that trading one investment for another with a higher yield (discount rate) can do nothing to meaningfully help meet a future obligation when the two investments have the same market value, because of the higher "risk" that goes with the higher yield must offset the expected benefit somehow, hence the equal market prices. And that this is thus "perpetual motion machinism", regarding both GM and SS.

If that were true, than the GM pension trust couldn't meaningfully reduce its risk of being unable to meet its future fixed obligations to retirees by investing in bonds and stocks rather than cash of equal market value.

I dunno, all this seems rather obvious and trivial, so maybe I'm misreading Sloan, and I'm not going to pursue it any more.

Jim,

OK. I'll quit too. But let me note we're using "risk" in two different ways. Your definition is "probability of failure," mine is "variability of return."

You are quite correct that the chance of meeting or exceeding a particular future target varies across portfolios. If the target is low, "cash" (short-term safe investments) gives the best chance. As it goes up you need higher risk investments.

And if investing were a binary matter, either you have $150X in ten years or the world ends, you would be right. But the world isn't going to end if you wind up with $140X instead.

Comments are Closed.