« Why Oh Why Can't We Have a Better Press Corps? (An Unusual Wall Street Journal News Pages Edition) | Main | Jonathan Rauch Loses It »

January 14, 2005

Risk and Administration Costs

John Corzine writes:

NJ.com: Search: Many privatization advocates rest their case on claims that seniors will enjoy better returns. However, such claims are misleading. First, they generally overlook the costs of financing the accounts -- the higher interest costs that future taxpayers will be forced to bear. Also, privatizers typically ignore the fact that Social Security, in addition to its role in protecting retirement security, also includes insurance for workers who become disabled and for survivors of workers who die prematurely.

Perhaps more fundamentally, privatization proponents generally fail to adjust projected returns for the added risk of investing in equities, as virtually all economists agree is necessary for a fair comparison.

Having earned my living as a trader and investment banker for 30 years, and having run one of America's largest financial companies, I understand something about markets. I can assure you it is pure folly to assume that privatized accounts will always increase in value and will be at a high-water mark at the moment when an individual retires. The truth is, markets go up, down and sideways -- sometimes for many years. One thing they never do is provide guaranteed returns or protection against both inflation and the risk of outliving your savings -- only Social Security does that.

There is another problem with privatized accounts: They are very costly to administer. One reason is that many accounts are quite small, so a significant share of any gains is eaten up by management fees. A University of Chicago study found that fees would reduce benefits by 20 percent. By contrast, Social Security's administrative costs are minimal, about one-half of one percent....

Posted by DeLong at January 14, 2005 12:22 PM

Comments

so why don't we look at the UK's problems which are those outlined here, and simply advertise it wide and far?

Posted by: Carol at January 14, 2005 12:31 PM


The admin says we need to reform social security because under a scenario of 1.7% real GDP growth
-- about half the long term norm -- the social security system will have a deficit. Thus people should invest in the stock market because over the long run stocks have provided average nominal long run returns of about 10%.

But, if growth is only going to be about half the long term norm it is quite probable that stock returns will also only be a fraction of their historic norms.

The internal economic arguments the admin makes are inconsistent.

Posted by: spencer at January 14, 2005 12:49 PM


"Many privatization advocates rest their case on claims that seniors will enjoy better returns However, such claims are misleading. First, they generally overlook the costs of financing the accounts ... Also, privatizers typically ignore the fact that Social Security ... also includes insurance for workers who become disabled ..."
~~~~~~~~~~~

I'm still waiting for the first Democratic defender of the status quo who tries to answer the "returns" issue with the "high cost" of private investments, who is honest enough to mention how -- starting just from now on, as a major historical change -- the SS status quo mandates growing **negative** returns for all future retirement cohorts.

As per the SS Actuaries. *Including* all non-retirement insurance benefits.

Still waiting ... waiting ...

Markets are risky -- but it takes a government to *guarantee* negative returns for 40 years.

Posted by: Jim Glass at January 14, 2005 12:58 PM


Even the past 10 years, the total market index has clipped along at better than 10% average per year. That includes both bull markets and bear markets. The stock market is the best place to be for the long term, and it's the point made in the book Stocks For The Long Run, by Jeremy Seigal (sp).

You bring up the issue of fees, yet many diversified investments (such as index funds and exchange traded funds) can be bought for under .18% per year.

I agree with you that we need to distinguish between the amount of money that goes to fund things other than retirement, such as disabilities. But certain some percentage that currently goes towards retirement, should be privatized.

Posted by: muckdog at January 14, 2005 01:05 PM


Senator Corzine also wrote a piece for Insight Magazine where he said that he was a strong proponent of expanding private accounts on top of Social Security. He said that he would add progressive measures, like tax credits, to help those who need it.

It looks as if the good alternative to the gutting of Social Security is emerging. Let's hope the Democrats get behind it.

Posted by: Brian at January 14, 2005 01:42 PM


muckdog wrote, "Even the past 10 years, the total market index has clipped along at better than 10% average per year. That includes both bull markets and bear markets. The stock market is the best place to be for the long term, and it's the point made in the book Stocks For The Long Run, by Jeremy Seigal (sp)."

There are two problems with this claim.

First, as Spencer points out, the claim is not consistent with the GDP estimates assumed by the Social Security trustees in their 75 year projection.

Second, stocks are somewhat more expensive now than their historical norm, and hence the odds that they'll perform well aren't all that good, barring an acceleration in GDP growth or the P/E ratio.

Posted by: liberal at January 14, 2005 02:00 PM


Jim Glass -

Well, Fox News always as a stable of "liberals" around to "admit" things which are not true, so maybe you should go troll there.

But perhaps you are confusing the current Social Security system with Bush's abolition-by-cost-of-living-indexing plan. Under the Bush plan, guaranteed payouts as a fraction of per capita income asymptotically approach zero. Now that's a negative return for you!

Posted by: aretino at January 14, 2005 02:06 PM


Jim Glass wrote,

"I'm still waiting for the first Democratic defender of the status quo...

"As per the SS Actuaries. *Including* all non-retirement insurance benefits."

Have a link for that? (Not that it would surprise me.)

"Still waiting ... waiting ..."

OK, if it's true, I agree to it, but only insofar as projections of future GDP the SS trustees are using are accurate.

The trick is, how to make up the difference, which is really the bill from the implicit debt of SS having been started as a pay-as-you-go scheme.

What's your answer to that one?

"Markets are risky -- but it takes a government to *guarantee* negative returns for 40 years."

But Social Security isn't an "investment," so the concept of "return" is only notional.

Besides, if you had accurate knowledge of the state of the world, you'd know there are many instances where the government has fantastic returns on its investments, and the private sector poor ones. Compare decades-long federal investment into research that lead to the discovery of erythropoietin. This allowed Amgen to come up with Epogen, but a careful reading shows that in total, Amgen's contribution was pretty minimal, and (in common fashion) Amgen is pretty much just a rent collector. And how good is the research they've done with all those billions in economic rents? Not very.

Posted by: liberal at January 14, 2005 03:17 PM


Liberal, overvalued compared to what? Can you name a rolling 20-yr period where stocks lost money? (Go to MSN's money site and give it a whirl.)

Posted by: muckdog at January 14, 2005 03:43 PM


muckdog,

Can you name a 20 year period where the economy grew at the 1.8% rate used by the Social Security Trustees to forecast future growth?

Also Social Security is not an investment, it's a guaranteed baseline of security.

Also how do we pay for the massive borrowing that would accompany privatization? (Over ten trillion for the first few decades of privatization)

Can you name a country that has privatized and been successful?

Posted by: KevinNYC at January 14, 2005 04:51 PM


Muckdog.

Why do you think a 0.18% fee is low? When people quote those kinds of numbers they are usually talking about 0.18% of the total stock, not 0.18% of the flow. If your investment returns 7% and the fee is 0.18% of the stock, then the net return will be 7% minus (0.18% / 7%) = 7% minus 2.6% = 4.4%.

Posted by: Mr. RDES at January 14, 2005 04:59 PM


All this reminds me of Mao's Great Leap Forward, where he asked every small town to have his own steel factory, operation that could double the production in no time. The fact that steel production needs a certain critical mass to be operationnal, could be easely overruled by the revoltionnary spirit. Someone remembers the outcome of this story ?

Posted by: FDK Brussels at January 15, 2005 02:12 AM


muckdog wrote, "Liberal, overvalued compared to what? Can you name a rolling 20-yr period where stocks lost money?"

Just because I said stocks are "overvalued" in some sense doesn't mean I implied they would lose money. The point about stocks being "overvalued" is that they very likely won't make the 7% annual real returns that they have in the past.

See especially KevinNYC's comment.

"Overvalued"---you haven't heard the oft-repeated point that the P/E ratio today is quite a bit higher than its historical norm? (Something like 20 versus 15 or so.)

Posted by: liberal at January 15, 2005 03:22 AM


Mr. RDS wrote, "If your investment returns 7% and the fee is 0.18% of the stock, then the net return will be 7% minus (0.18% / 7%) = 7% minus 2.6% = 4.4%."

???

Posted by: liberal at January 15, 2005 03:35 AM


I believe Jim Glass is quoting a negative return compared with a benchmark. Perhaps against the treasury bond yield. Don't know if things like disability and survivor benefits are included in that calculation. 2.4% out of the 12.4% of payroll taxes currently go to the disability trust, where, presumably, most people are a tad more concerned about insurance than return on investment.

As for negative returns on equities in any 20 year period - I recall reading that the worst historically has been 3% in a 20 year period. Of course, that is before adjusting for inflation. So, there'd be historical periods during our narrow window of history when the economy always grows, when equities have a negative real return, and a big negative return compared with the same benchmark that I believe J Glass is using to project negative returns for all future SS recipients.

A big falacy is extrapolating forward from historical data. In the parlance of disclosure, "Past returns do not guarantee future results". One simple measure is to note that, since the mid-1970s, we've had an annual tailwind of about 2.7% because people have been willing to pay more and more for a given level of earnings. If price/earnings levels were to revert to historical norms any time during the next 30 years, that'd amount to a headwind of about 2.7% per year for equities.

Another big falacy is that average returns don't show how individuals fare. We could well have a situation where the "average" return over an extended period exceeds the treasury bond yield, but the "median" return doesn't.

Posted by: Charlie at January 15, 2005 04:00 AM


Liberal asks what I'm talking about. I'm pointing to the old distinction between stocks and flows. You have to be careful when you read about the transaction costs to manage private accounts. Very often these kinds of apparently low values (such as 0.18%) are actually referring to 0.18% of the entire account, or stock. If the return, or flow, from that stock is 7%, then that 0.18% charge against the stock account would take off roughly 2.6 percentage points off that 7% return. So if the 0.18% number means 0.18% against the flow, then that's a good deal. If it means 0.18% against the stock, then it's not.

Posted by: Mr RDES at January 15, 2005 06:02 AM


Mr RDES wrote, "Very often these kinds of apparently low values (such as 0.18%) are actually referring to 0.18% of the entire account, or stock."

Of course it's 0.18% of the entire account. That's usually how operating expenses are reported, at least for mutual funds.

"If the return, or flow, from that stock is 7%, then that 0.18% charge against the stock account would take off roughly 2.6 percentage points off that 7% return."

Nonsense. The appropriate computation is clearly 7% - 0.18% = 6.82%, not your completely incorrect "7% minus (0.18% / 7%) = 7% minus 2.6% = 4.4%."

Posted by: liberal at January 15, 2005 08:10 AM


If you have a 100 dollar portfolio with a 7% return at the end of 1 year it is worth $107.

The annual fee will be 0.18% of the $107 portfolio, or $0.19. Subtract that from the $107
to yield $106.8.

That is how the fee reduces the return.

Posted by: spencer at January 15, 2005 08:52 AM



Hey! My auto insurance premiums have been generating a strongly negative return, and the numbers show that I'm far from alone.

I'm ready to join with all the Bush-worshiping economists and demand that I be allowed to invest my premium payments in the stock market. After all, if you look at the average result, everybody wins.

Through the magic of the market, I will never have an accident that costs me more than the balance in my priva...er, personal...Auto Liability Investment Account.

Posted by: Ottnott at January 15, 2005 03:01 PM


Ottnot,
My family is self insured for flood insurance. We were flooded, slighly, only $14,000, in the last el nino years so we looked at what flood insurance costs and what our personal flood wall would cost, about $5000 and decided to drop the $1000 per year flood insurance. We of course invest the extra $1000 per year in disposable.

A socailized approach is being implimented for the flood risk but will take more time than we have lifetime.

We also do the same with earthquake insurance. You get more value earthquake proofing than insuring.

Some risks are something that can be delt with on the individual level and some have to be socialized. We pay taxes for the socialized risks and also volunteer with the county sheriffs office disaster assistant teams.

Others milage may vary.
Enjoy

Posted by: dilbert dogbert at January 16, 2005 09:14 AM