Across the Curve is gone.

Accrued Interest is gone.

I’m still here, but bonds aren’t my sole focus. Long time readers know that I have a wide variety of interests in investing. I sometimes think my following is smaller, because not everything I write appeals to my audience. It is hard to predict what I will write about.

Are there other good bond blogs? I haven’t seen any, but maybe my perspective is limited.

This post is a call to all who think they have something to say about the fixed income markets — do you want to try to create the definitive fixed income blog? I would be willing to host and moderate it, but I would need bright, aggressive minds to write and explain to readers what is going on in fixed income.

Why?

The bond market is bigger than the stock market, and those that invest there are brighter in one sense — they have to make decisions over small differences yields, versus the safety of those yields. They are dumber in another sense, because the rewards of managing equities are larger than that of bonds except at the largest managers.

So, I put out the call to those that want to write about the bond market. Do you want to create one great blog? Let me know, and maybe we can make a difference for those that need to learn about fixed income. Remember, this is about giving something back. We are paid well enough, but can we create something useful for the broader investing public?

I have no idea as to whether this could work or not, but I welcome your thoughts.

## 42 Comments

There is also the bond vigilantes blog from M&G here in the UK:

http://www.bondvigilantes.co.uk/blog/

David — I like the idea, but immediately have two questions:

1) Are you thinking one person write this new blog? I got the impression that John Jansen and ??? (I can’t remember the Accrued Interest author’s name?) had trouble keeping up by themselves — not to mention both authors had more expertise in particular sections of the bond market… multiple authors would have less work (each) and might have complimentary areas of expertise (Treasury, MBS pools/TBA/CMOs, corporate, futures, etc)

2) Who is the audience for this blog? Retail gets smashed buying bonds for anything other than buy & hold (very wide bid/ask). Retail investors almost have to buy funds or else a ladder of buy and hold. Trading bonds for retail mean futures (for the brave) and/or ETFs (or getting “your face ripped off”).

Institutional investors (and retail?) would benefit from some decent analytics that are “independent” (i.e. not supplied by a sell side firm) … not to mention those analytics give a good jumping off point for blogging / comments (doesn’t matter if you agree or disagree with the analytics, so long as you explain why)

I might be interested in **helping** with a bond blog and might know a couple others … but taking on such a task single handed would be a huge undertaking

As a member of the broader investing public I use your blog to educate myself.There is very little in

the blogging or library universe an average person

can use regarding all manner of bonds.Specifically,

the advantages and detriments to using large mutual

funds of bonds as a proxy to purchasing a contractual bond.I read Fabozzi 7th edition on fixed investments as far as I could.

If your concept bears fruit one of the goals should be to lift the general public above Bonds 101 and help them navigate head fakes and asshattery.

Good luck

Fred

For that matter, can anybody explain why the main financial news sources report so little on the credit market relative to the equity market? Mark Cuban raised this question a couple years ago, if I remember right. No answers. Strange. It can’t be just the explanation of any earlier commenter — ie, that there’s little opportunity for retail. Since when has that stopped reporters from covering IPOs or private equity.

Michael Martin — ” why the main financial news sources report so little on the credit market relative to the equity market?”

Another hesitation I would have (that I didn’t mention earlier) concerns retail investor education.

I used to work at a sell side bond shop, and my department did presentations for institutional clients.

When I presented to US based clients, I would go through a very lengthy series of graphs and algebra to explain many concepts. It would take 30-40 minutes for each one.

When I presented to Asian clients, I would just say “its the first derivative of this equation”. The clients “got it” in 2-3 minutes instead of 30-40 minutes.

Bonds are all about math. Calculus isn’t “required”, but it often gives you a big leg up. If you want to look at yield curves, its 100x easier if you can understand matrices. Bond option pricing is all about statistics. Taylor series are all over the place.

While there are many many many Americans who are good at math, the unfortunate reality is that the vast majority are not. How many people know how to compute the interest rate on their passbook savings account?

People in the US who are good at math are labeled nerds and looked down upon. Presidents (including but not limited to Obama) have often commented on the lack of science / math majors in our society.

Our society values sales ability above all else, often to the exclusion of all else. Hence, CNBC has people shouting over sound effects and fancy computer graphics. Some big boobed blonde smiles and tells you how much your stocks fell today! When was the last time you saw any math equations on CNBC? When was the last time you saw any math on any widely watched channel?

Almost every day, I see a talking head say that IRAs are advantageous because they let you defer taxes … WRONG!!! Anyone who understands the distributive property of multiplication should immediately see why.

If you take any series of returns, and you subtract out taxes each year — and then compare that to the exact same annual returns and pay taxes once at the end — you get exactly the same outcome. EXACTLY. Its not rocket science, its the distributive property of multiplication that your elementary school teacher bored you silly with.

IRAs make sense only **IF** you will be paying a lower tax rate when you are older (when you take withdrawals). In hindsight, that turned out to be true for our grandparents. With the way our government spends money today? Its entirely possible that tax rates will be higher when people start to take money from their IRA

That is just one example that effects millions of people. Is the mass media deliberately lying to its viewers? Or, are the talking heads on TV all too representative of the audience?

Greg:

“If you take any series of returns, and you subtract out taxes each year — and then compare that to the exact same annual returns and pay taxes once at the end — you get exactly the same outcome.”

That’s only correct if all of the following are true:

1) All of your capital gains are deferred to the draw-down date

2) Your securities do not produce income or otherwise make distributions

3) The tax rates on capital gains, dividends, and income are the same

David:

I would suggest that a large majority of people who are experts in credit markets are muzzled by employers’ codes of conduct. The company at which I work even has a mandate against… commenting on financial blogs, even on our off-hours (like I am doing right now). I consider the policy to be unconscionable, so I ignore it; however, generating high quality content is a lot more visible (and ethically grey).

Matt –

#1 is true for IRAs

#2 is not a condition at all — whatever income you get is included in your annual returns. You get the income either way

#3 the tax rate (singular) is the same for IRA withdrawals, regardless of whether it is cap g/l, dividends or income

IRAs have no advantage at all unless you expect to have a lower tax rate when you start making withdrawals

oops — you are right on #3…

If current tax rates for cap g/l dividends are less than for regular income (as is the case at the moment) … IRAs are actually WORSE than a taxable account

david

this has been a good source for me on the bond market, but I fear the author will not be continuing it with his new employment:

http://acrossthecurve.com/

ps personally, I hope you continue Aleph Blog-one of my ‘favourites of favourites’. Thank you, David!

The best site I’ve used for bond news is BondHead weblines at http://www.bondheads.com

It has a compilation of news on bonds, economics, and related subjects that can be helpful. It’s updated once a day, so while it is of limited use for up to the minute market-moving news, it effectively surveys current news. Possibly the folks who run that site might be inclined to help fill the void.

Matt — the reason most compliance departments will not allow you (us) to comment on financial blogs is because they fear readers (and their attorneys) will “re-interpret” whatever you write.

That many people have math “misunderstandings” means that anything you write (with an assumption that people understand math) can and will be misinterpreted according to people’s misunderstandings.

And then you find yourself trying to explain calculus to 12 people on a jury who didn’t take calculus in school, and certainly don’t want to hear about it on jury duty….

It isn’t just math. I have experienced several cases on *this* blog where it is obvious the other commenter has reading comprehension problems. You might disagree with me and whether my writing was unclear / that other person has no reading comprehension — but LOTS of college professors, the SAT testing company, and employers have complained about lack of reading comprehension skills

If readers have trouble understanding the written word, and are lost with the math — you have a huge potential legal liability

Remember, this is a country where a lady won a legal award because the microwave manufacturer didn’t say not to put Fluffy the Cat in the microwave to dry.

“the rewards of managing equities are larger than that of bonds except at the largest managers”… hmm… this betrays your focus — on bond funds and bond fund managers — who are by and large an insipid lot producing ho-hum returns; whatever they talk and blog about can perhaps make a “great blog” (for them) but it sure won’t make an exciting one (for the rest of us). Which is too bad because bonds really can deliver a punch — returns on bank sub debt in 2009 were on average in excess of 300%. but to be able to blog about that you have to target your blog to some other demographic than bond fund managers.

“the rewards of managing equities are larger than that of bonds except at the largest managers”

I skipped over this comment yesterday. David himself has argued (in other posts) that the so called “equity premium” is a measurement error (time weighting vs capital weighting).

The other thing this comment omits is leverage. Most bond funds are unlevered (a few are, but most are not). Equity funds (even when not explicitly levered) have embedded leverage — the companies themselves have assets far in excess of shareholder equity.

If you control for that leverage, bond and equity mutual funds have very similar results (on average)

David

You state: Almost every day, I see a talking head say that IRAs are advantageous because they let you defer taxes … WRONG!!! . . . If you take any series of returns, and you subtract out taxes each year — and then compare that to the exact same annual returns and pay taxes once at the end — you get exactly the same outcome. EXACTLY.

Sorry, but no: Assume thirty years of IRA contributions at $2,000 per year made at the beginning of the year, with consistent earnings of 15% per annum. If the earnings are taxed annually at 35%, the ending value is $344,000. If earnings are tax deferred, and taxed at the end of 30 years at 35%, the ending value is $671,000. I stopped my math education at calculus, but $671K is not EXACTLY the same as $344K, but is rather more (some might say materially more.) Perhaps you should focus less on the distributive property of multiplication and more on the beauty of compound interest.

Greg: “Almost every day, I see a talking head say that IRAs are advantageous because they let you defer taxes … WRONG!!! Anyone who understands the distributive property of multiplication should immediately see why.”

The deferred taxes are a benefit because the deferred sum can also compound.

Compare amount X which returns 10% per year for 30 years, and then pays taxes at 35%:

X (1.1)^30 = 17.45X, then (.65) = 11.34X

To same amount X which returns 10% per year, and is taxed each year for 30 years at 35%:

X (1.065)^30 = 6.61X

Wide Moat — you were supposed to learn about the distributive property of multiplication before graduating high school. And you were supposed to learn how to properly solve world problems also. 0 for 2.

You have successfully parroted the sales pitch / what the teacher told you to think. You graduate with an “A”!!!

Thank you for proving my point that most people are mathematically illiterate.

Let X = (1.1 * 1.05 * 1.07 …) or make it (1.1)^30. Whatever you want to assume your return series to be

Now you want to argue that

0.65 * (X)

is not equal to

(0.65 * X)

Or are you a member of Congress and proposing that taxes should also be compounded?

After learning the distributive property, some learn the standard order of operations.

(.65) x^y does not equal (.65x)^y

Wide Moat — Yes, you should learn the standard order of operations — exponents do not have any higher priority than multiplication/division (exponents are essentially multiplication, duh!!!)

Deferring taxes makes sense if you expect to be in a lower tax bracket when you pay, and it is no different with pensions. paying your taxes this year, or paying extra taxes next year on (1*r) is the same.

IRAs offer the same “tax deferred” advantage as a whole life insurance policy, which was (and is) available all along. We just can’t draw any general conclusion about life insurance commissions versus IRA investment fees/commissions — so can’t say what is better or worse ( and of course you have to back out the insurance cost in the policy since the IRA doesn’t offer that)

.

@@David Merkel: I hate to admit the lawyers are right, but… the legal liability of writing for a bond blog is way too high

Greg, I don’t mean to put you on the spot but Wide Moat is correct. You are mistaken in thinking that exponents do not have precedence over multiplication/division. In fact, exponents and square roots have the highest precedence of all (ignoring brackets–brackets actually have the highest precedence but they are not a math operation per se). HEre’s a wikipedia entry on it:

http://en.wikipedia.org/wiki/Order_of_operations

Anyway, to bring it back into investing… the tax deferral is one reason Warren Buffett doesn’t pay dividends (the primary reasons are actually the double-taxation of dividends and the ability for him to re-invest at high returns (although that, arguably, hasn’t been the case for years) but the tax deferral is also a key reason.)

Michael Martin: “For that matter, can anybody explain why the main financial news sources report so little on the credit market relative to the equity market? “I think it’s because stocks, in general, have done very well in the last 30 years. Sure, there have been some big crashes of late (2000 and 2008) but stocks have clearly captured the interest of the masses (not to mention fund managers–many institutional funds were going into the recent bust with near historical lows of holdings of bonds.)

As for the argument by some that bonds are complicated to analyze, that’s true only to some degree. Yes, some of them require more mathematics and quantitative thinking; but I wouldn’t say they are more complicated than stocks.

In fact, the most difficult asset to analyze is probably common stocks. I’m probably the dumbest one around but even the smartest and most successful of you would probably come closer to correctly valuing a bond (whether it’s a govt bond or a corporate junk bond or an ABS) than you would in figuring out how much Coca-Cola shares are worth or how the share price of Kodak should be?

Having said all that, bonds are not accessible to small investors. As long as that remains true, I think it will remain the playground of the professionals and institutional funds. I can’t see too much interest in a bond blog, other than from professionals or academics.

sivaram velauthapillai — we can argue about order of math operations all day (I think you are mistaken), but there are two problems here:

1) the decision to write (x * x * x …) or (X^30) is quite arbitrary, so explain why you think exponents should be more important in this example? In real life (not theory), you will never get the exact same return every year so the problem should be expressed as a series of multiplication problems.

Also, anyone familiar with compounding knows the order of the multiplication series is not only relevant, it is critical. The series is not commutative.

Hence, order of operations isn’t even relevant to this problem

2) Second and more important, Wide Moat is modeling the problem incorrectly — which is why he is getting a wrong answer and why he doesn’t see the distributive property factor

As for Warren Buffet’s dividend policy — he has argued about double taxation and his reinvestment abilities many times. I have never heard him argue about tax deferral as justification.

It is true that constantly selling investments (to pay cash dividends) creates tax problems versus buy and hold (is this what you mean by tax deferral?) Paying dividends isn’t the tax problem — its selling investments to pay the dividends that creates a tax liability. Buying and holding the investments (instead of trading) allows tax deferral — but the tax liability is the same.

If he wants to actually reduce his tax bill, he has to donate his money to Bill Gate’s foundation

Wow, the irony has me almost laughing out loud.

Greg, you seem a nice, well-intentioned young man. A couple hints for you: 1) when you find yourself in a hole, stop digging 2) never argue algebra with an Indian.

My advice, you might want to get yourself a new handle.

AllanF — one what basis do you assume I am a “young man”? Obviously, you intended the comment as an insult and a put down since you have no basis for guessing my age at all.

Indians do not have any monopoly on math skills– on the other hand Indian call/help centers are notorious for claiming english skills that they simply don’t have. Siviram doesn’t understand the basic problem — and instead wants to argue math minutea

0.65*X*X*X is exactly the same as X*X*X*0.65

and if you phrase the IRA problem incorrectly (as Wide Moat did) — that point is not disputable.

Integer exponents are a notation convention, not a math operator at all. Real exponents are a math FUNCTION, which gets higher priority than parenthesis — but that isn’t the example Wide Moat gave.

The actual problem is that Wide Moat did not set up the problem correctly — which is why he got the wrong answer, and also why an Indian guy couldn’t understand the garbled English.

The actual question of expected value at the end of a cashflow stream involves compounding — that is to say the order of the returns are critical. You need to maximize the number of years each cashflow can compound. A “bust” early in the process means the rest of the equation goes to zero (as does the tax liability) — unless you have additional cashflows / deposits.

Once you look at the problem as a time weighted problem, you will see the distribution issue I brought up many comments ago.

And AllanF — my advice to you is not to insult people based on their disagreeing with your biased preconceptions. I would also encourage you to be less of a racist

Sivaram: “the tax deferral is one reason Warren Buffett doesn’t pay dividends (the primary reasons are actually the double-taxation of dividends and the ability for him to re-invest at high returns (although that, arguably, hasn’t been the case for years) but the tax deferral is also a key reason.)”

Here’s one passage where Buffett explains this:

“Because of the way the tax law works, the Rip Van Winkle style of investing that we favor – if successful – has an important mathematical edge over a more frenzied approach. Let’s look at an extreme comparison.

Imagine that Berkshire had only $1, which we put in a security that doubled by yearend and was then sold. Imagine further that we used the after-tax proceeds to repeat this process in each of the next 19 years, scoring a double each time.

At the end of the 20 years, the 34% capital gains tax that we would have paid on the profits from each sale would have delivered about $13,000 to the government and we would be left with about $25,250. Not bad. If, however, we made a single

fantastic investment that itself doubled 20 times during the 20 years, our dollar would grow to $1,048,576. Were we then to cash out, we would pay a 34% tax of roughly $356,500 and be left with about $692,000.

The sole reason for this staggering difference in results would be the timing of tax payments. Interestingly, the government would gain from Scenario 2 in exactly the same 27:1

ratio as we – taking in taxes of $356,500 vs. $13,000 – though, admittedly, it would have to wait for its money.”

http://www.berkshirehathaway.com/letters/1989.html

Inputting Buffett’s numbers into my formulas above (comment #15), and compare:

a) 1 * (2)^20 * .66 = 692060

[1 = initial amount, 2 = rate of return, 20 = years, .66 is remainder after 34% tax]

b) 1 * (1.66)^20 = 25244

[1 = initial amount, 1.66 = annual rate of return after tax, 20 = years]

You may pay 34% tax either way, but in the latter case, you have 27x more money.

As you know Sivaram, therein lies the most profound defense of buying businesses with wide economic moats that can also reinvest capital at a high rate of return.

OK — trading day is over so I can actually put my original statement on IRA taxation / final value to the correct math equation (not the erroneous one Wide Moat gave).

Assume:

A=annual payment to your IRA ($3000 or whatever)

T=(1-tax rate) … with 35% tax rate = 0.65

r=(1+annual return) … stick with 10% = 1.1

…and lets further assume 30yrs to retirement

In the simple one year contribution case:

_ IRA = A * r^30 * T

or

_ Taxable = A * T * r^30

Do the math for yourself, you get $34,026.33 both ways. “Wide Moat” assumed in his equation that taxes get compounded, which is incorrect.

Now look at the more complex case for the taxable account, where contributions are made each year:

_ Taxable = ((A*r^30*T) + (A*r^29*T) + (A*r^28*T) … (A*r*T))

now using that annoying DISTRIBUTIVE property thing, pull out all the terms A*T, and you get:

_ taxable = A*T*((r^30) + (r^29) + (r^28) … (r))

And once again, it does not matter if you put the tax term “T” in the front or the back of the equation, you get the same numerical answer. Put it on the front of the equation, you have a taxable account; put it at the back, you have an IRA. But the numerical value is the same.

If you want to assume each year gets a different rate of return and/or a different contribution– then you can only divide out the “T” term. But if you make the same return & contribution assumptions for both taxable and IRA accounts, you get the same numerical result at the end.

If is only the case where you assume a better (lower) tax rate when you actually pay the taxes (now or later) that you benefit.

The elderly generally have lower incomes (and thus lower tax rates), which is why a retirement account made sense in the past when IRAs weren’t available.

For current workers, the question is your tax rate today versus the one you expect to be paying when you retire. Given Washington’s spendthrift habits there is every possibility that tax rates in the future could be much higher than today, rendering IRAs a bad choice compared to saving in a taxable account (not to mention lost flexibility).

For those of you that still think A*T*(r^x) is has a different value than A*(r^x)*T … well the government is selling lottery tickets to all math wizards, and you just might win!!!

Wide Moat — you are still (comment #25) assuming that taxes get compounded. That is wrong.

You cannot take your 35% out of the return number and then raise it to whatever power — that causes the taxes to get compounded. Taxes do not compound

Buffet’s example (that you cited in comment #25) is the answer to a totally different question.

This is an investment decision (selling every year, paying tax, and then reinvesting vs buy and hold that Buffet advocates). That is a different question from the investment vehicle used (IRA versus taxable)

The tax deferring “advantage” Buffet cites is the advantage of buy and hold over constant trading — not the advantage of executing whichever strategy

inside various legal entities.

Warren Buffet’s investment vehicle / legal entity is not a tax deferred IRA — it is a corporation that pays taxes each and every year.

His “tax deferral” comes from his buy & hold investment style.

If you “buy businesses with wide economic moat” and hold them forever, the IRA investment vehicle affords you zero advantage

Stop digging.

AllanF: Stop your racist comments and learn some math skills

@ David;

15 of 31 comments on this thread are from Greg; including 4 in a row today.

Perhaps he should be encouraged to start his OWN blog?

He certainly has the time and inclination …

@ Lurker — agreed. Guys, let’s drop the topic. Greg isn’t likely to be convinced.

Dave,

All the best in this endeavor.

John Jansen

I have to come to Greg’s defense here. The result is the same. Try it will $100 for 3 years, 35% tax rate and 10% per year – you will come to 236.66 either way

Greg, the calculation you keep describing is basically the comparison of regular IRA vs. ROTH IRA – e.g. if ‘C0′ is the starting amount, T1 is the current tax rate, T2 is the retirement tax rate, and x is the compounding factor (taken to be constant for simplicity – e.g. if the return is 10% then x is the constant 1.1), then

1) Traditional IRA over n years followed by a lump sump payout looks like C*(x^n)*(1-T2)

2) Roth looks like C*(1-T1)*(x^n)

3) Taxable account with all non tax-efficient gains looks like C*(1-T1)*(1+(x-1)*(1-T1))^n =

C*(1-T1)*(1+x-T1x -1 +T1)^n =

C*(1-T1)(x -(x-1)*T1)^n

By the assumptions used, x-(x-1)*T1 is less than x.

The benefit of the IRA, given the simplifying assumptions used, is that it compounds at the rate of (x-1) instead of the lower rate of (x-1)*(1-T1).

Josh has finally stated the right explanation – at equal tax rates, Traditional and Roth IRAs are equivalent. Paying taxes every year rather that all at the end is in fact detrimental to compounding because current dollars are worth more than future dollars (assuming some level of inflation/positive returns). Amazing that someone who holds such an attitude of superiority and condescension (Greg) has posted 15 times without realizing this basic fact.

Another way for the slow to understand this without resort to calculus: is the Roth IRA provision beneficial to the taxpayer? Yes. Tax-free gains are obviously greater than taxed gains. Is the traditional IRA equivalent to the Roth IRA at equal tax rates based on the distributive property? Yes. Therefore the tax deferral on contributions and gains has value.

So after reading Josh and najdorf’s comments, they both think I phrased my comments in a condescending way — but they both agree with me.

At equal tax rates, there is no advantage / difference between paying taxes now or later — which is what I said from the beginning.

It is important to note that BRK/Buffet (and plenty of others) gets the exact same “tax deferral” benefit WITHOUT an IRA — the IRA doesn’t matter.

Sorry to David Merkel for monopolizing his blog on this, but the message from other commenters is crystal clear: you are not allowed to post comments that disagree with the crowd’s view.

I had the same problem at a sell side bank– saying subprime loans were going to ruin us. I basically got run out in a power play that I lost, but today that bank is gone. Unpopular views aren’t always wrong

And as David point’s out in his newest post — there is no monetary reward in sharing even neutral viewpoints, never mind controversial opinions. This is why the country has crappy politicians, and its why “the crowd” happily leveraged themselves up to oblivion and then complained that the banks made them do it.

Our enemies are all around us — in every mirror in every room

I am late to the party, but Josh is right to explain that there are two effects here: commutativity of multiplication and exponentiation of factors.

David: I am very much interested in the bond market as a retail investor. I don’t care about real time market commentary in as much detail as John Jansen did – once a week summary would be fine. I don’t know how I would contribute though, as I know a bit of bond math (I am an engineer with PhD in CS, should probably be working for one of the bucket shops after all what I am reading), but only see the trading details as an outsider and interpret what I read on the news and blogs.

There is another group of people interested in bonds at https://self-evident.org

Nemo did some explaining of JJJ jargon and math, while Bond Girl had her own blog at some point focused on states and munis. I never comment on self-evident as it requires registration, but maybe you should invite them to the project. I very much would appreciate any efforts in that direction.

Greg, couple points. First, I didn’t make any remarks about your style. Just attempted to clarify the math. Second, your attempt at a summary immediately above is incorrect. What is correct is that if someone bought ETFs in a taxable account that paid no dividends and they never sold those ETFs until retirement, then *that* would completely equivalent to holding the same ETFs in a Roth IRA. However, in the more realistic case where they are paying taxes ondividends and capital gains (either long or short term) at various points prior to retirement, then their overall return would be *lower* than holding the same investments in a Roth IRA, since they receive no value of the later year investment returns on the monies that were subtracted from the account in order to pay taxes in the earlier years. The formulas I gave above presented this for the special case where they pay taxes on all gains at the short term rate in every year. Obviously, you can rework the example to account for intermediate levels of tax efficiency. But the returns in the taxable account only reach the level of the Roth returns in the extreme, unrealistic case, where there are no dividends and no intervening capital gains in the taxable account.

The summary above is at about the limit of how simply I can state the issue, so I hope it does the job. In the bigger picture, I am sympathetic to your POV that the media often gives out dubious financial advice as received wisdom. In the case of IRAs, their largest benefit for the average punter is not the tax efficiency but rather the disciplined savings pattern that they represent. Likewise, for the advanced investor, their drawbacks in the form of higher average transaction costs compared to maintaining all funds in a single taxable account is IMO not insignificant, though harder to analyze.

I will mostly ignore Greg’s last post because he seems to have changed his stance to the opposite of his original point – now he agrees with Josh and I and believes that it is rational for BRK to defer taxes and that an IRA can give investors in any sort of asset the same advantage.

On the bond blog point, one area that I think is under-explored is bankruptcy/default situations. Typically you see a lot of shallow media coverage of trading in floundering companies, then they disappear into a black hole. For instance, can anyone readily find me a list of the largest corporate or muni bankruptcies of last year on a free source? Can anyone find good resources on recoveries? I assume Bloomberg or S&P have this info behind a paywall, but bonds are typically sold to retail on the premise “Hey, it yields 6%!” with very little analysis of default probability/severity. Media coverage continues the yield focus with very little attention to final economic realizations. We’ve seen some very interesting restructurings this year but I’ve been unable to find in-depth information about precisely what bondholders got and how the negotiations worked.

Likewise for the endgame of some of the toxic waste MBS – in many cases the issuers stopped reporting to the SEC because the number of holders was so low, but if the holders are publicly-traded banks then I would like to be able to look at the quality of the assets in order to assess bank balance sheets and macroeconomic realities. I think the SEC should be looking a lot harder at requiring companies to report what debt they hold and how it’s performing – currently I look at a lot of bank balance sheets and see billions in assets with no clear risk metrics or information about counterparties. Knowing that current NPAs are 4.2% and that the bank has guessed reserves of 3.5% will be adequate doesn’t really allow me to make an investment decision. Why shouldn’t I, the owner of JPM, be able to see how much credit the managers have extended to various corporations and the precise makeup of their MBS portfolio?

The analyses of 401k vs Roth vs taxable account assume that financial assets will be taxed only once. Perhaps I am too skeptical of our government, but I can easily imagine a future in which the Roth IRA that I paid taxes on in 2008 will be taxed AGAIN in the future when I begin taking distributions.

If one believes this, then I think deferring taxes as long as possible is a rational approach.