The other day I was reading through the Pabrai Investment
Funds Annual report for last year when it suddenly occurred to me that the
comparative returns that Pabrai publishes there (i.e. his funds’ returns versus
the various indexes) is misleading. Why you ask? Well, it all has to do with
taxation.
Let’s have a look at the Pabrai investment funds:
Pabrai Investment
Fund 2 (an Illinois Limited Partnership) – Commenced operation on
1st October 2000 when units were issued at $10 each. As at 30 June
2012, the net asset value (NAV) of a unit was $43.36, an average annual
compound return per annum of 13.3%.
Pabrai Investment
Fund 3 Ltd (a British Virgin Islands corporation) – Commenced
operation on 2nd February 2002 when shares were issued at $10 each.
As at 31st December 2012, the NAV of these shares was $33.21, an
average annual compound return per annum of 9.7%.
Pabrai Investment
Fund 4 (a Delaware Limited Partnership) – Commenced operation on
1st October 2003 when units were issues at $10 each (Pabrai must
have a fondness for the figure of $10). As at 31st December 2012,
the NAV of these shares was $18.85, an average annual compound return per annum
of 7.1%.
Now, the British Virgin Islands Corporation (Pabrai Investment
Fund 3 Ltd) pays no income tax or capital gains tax in the British Virgin
Islands and also pays no income tax in the United States. This is the reason
that Pabrai domiciled this company in that jurisdiction. It’s also a fund for
non-US investors. There is nothing wrong with this, it’s good tax planning.
But when Pabrai compares the returns of Pabrai Investment
Fund 3 to the returns of the Dow Jones Industrial Average (DJIA), the NASDAQ or
the S&P500, he is comparing the returns of a fund which pays no tax with an
index whose constituents must pay tax.
For a moment, imagine that the US Federal and State
Governments legislated to exempt all constituents of the DJIA from paying tax.
What affect would that have? It would immediately increase the value of all the
constituents of that index. Rationally, the percentage increase would be by
approximately the tax rate divided by one less the tax rate multiplied by 100
or (TR% / (1-TR%) )*100. Of course, it’s not quite that simple, but you get the
gist of it.
When Warren Buffett (or any other person running a US
domiciled tax-paying investment corporation) compares their corporation’s
returns to an index, it’s a fair comparison because like is being compared with
like.
When an individual running a non-tax paying corporation
compares their returns to an index which consists of tax-paying corporations,
the comparison is not fair. A fair comparison would be to take the
returns generated by the non-tax paying entity and then apply a notional tax
rate based on the average tax rate paid by the constituents of the index being
compared to and then compare the returns.
For example, if we applied a notional 25% “tax” to Pabrai
Investment Fund 3, that 9.7% average return (mentioned above) might become
about 7.3% - hardly returns to write home about. Incidentally, Pabrai
Investment Fund 3 is the only vehicle currently open to new investors.
But what about Pabrai’s limited partnerships? They too pay
no tax. All tax is paid by the individual unit holders.
So I would put it to you that any performance comparisons
between any of the Pabrai funds and the various US indexes are misleading
because any reported out-performance will always be skewed unfairly in Pabrai’s
favour.
Pabrai has talked at some length in the annual letter about
how he has compounded his own net worth at close to 26% per annum (since about
1995). That’s great for Pabrai, but the funds mentioned above haven’t achieved
anywhere near this figure and as I’ve explained here, even those returns are
misleading when compared to any index. So why mention your own much better
individual performance?
My own net worth has increased at an average annual rate of
16.1% (post tax) from 1995 to 2012. On a pre-tax basis (if I had never had to
pay income or capital gains taxes) it would have been roughly 23% per annum.
That statistic alone demonstrates the difference between pre-tax and post-tax
returns.
So, finally, to have some fun, if I had issued shares at
$10 each (Pabrai’s favourite number) on 31 December 1995 in “The Stock Scribe”
(me), as at 31 December 2012, those $10 shares were worth $127.
thanks for the post, just read the same docs from Prabai Funds also and didn't realise this.
ReplyDeleteInteresting article.
ReplyDeleteBit confused on a few points:
(a) Given the absense of franking credits in the USA and the fact that Pabrai's portfolio is also, I presume, consistent of tax-paying corporations, isn't an investor in his funds also investing in tax-paying corporations too, just like an index? I.e. dividends (paid after deducting interest and tax from earnings) flow through to unit holders whereby they are taxed at an individual's tax rate in both cases
(b)Since you deem a 9.7% tax exempt annualised return to be obscuring reality by the use of an alternative investment structure, then how can you deem Berkshire's comparison to an index any more reasonable? Leaving aside the proxy of book-value to intrinsic value, surely a corporation who uses low-cost insurance "float" to purchase private businesses (with less than 1/3rd of NAV/Equity used in actual stock market investments) comparing itself to an index which is 100% in stocks is no less obstructive?
Thanks in advance!
Your first point is correct. Pabrai's investors get taxed the same way regardless.
ReplyDeleteThe main point I'm making is that when Pabrai says a certain return was achieved, it's always pre-tax. A company like Berkshire pays tax, so that proportion of its returns has already been paid to the IRS. If Berkshire paid dividends its US shareholders would be taxed again as they are on the capital gains, so there are 3 sets of taxes being paid here: corporate income tax and personal income and capital gains tax in the hands of the individual. Pabrai's returns are only taxed in the hands of the individual.
You should note that Berkshire's insurance float is only low cost due to the extremely skilled personnel managing it at Berkshire. There is no guarantee that this float will always be low cost.
But more importantly, tax has to be paid on insurance profits and the private businesses purchased also pay tax. All of which makes comparing Berkshire's returns to an index much more appropriate than comparing Pabrai's returns in the same manner.
Like my comment below, the problem I have with this explanation is that you are essentially using two different lens when looking into the index and the performance of Pabrai funds.
DeleteIn fact, if we imagine S&P500 becomes S&P1, and the Pabrai Funds holding that exact same stock, the whole thing should be much clearer here.
Yes, in fact, they should have the same performance because they track the same stock. The only difference will be the management fees and performance fees Pabrai is going to take. But an investor in the index and another investor in the Pabrai Funds will have the same tax situation.
Both Berkshire and companies in the DJI (and S&P 500) are US domiciled and pay US corporate taxes (lets give Tim Cook the benefit of doubt). Thus, the net performance after tax makes a fair comparison.
ReplyDeleteWho is auditing Mr Pabrai Mumbai's returns? Probably some ex-partners from Arthur Anderson who enjoy the strict accounting standards in the British Virgin Islands more than Houston, Texas :-)
D
Thanks for the response! Great Blog!
ReplyDeleteP.s. Good call re:Mathews Capital given whats happened since.
ReplyDeleteP.s. Good call re:Mathews Capital given whats happened since.
ReplyDeleteWhat has "happened since"? Has Phil decided to do something different?
1. Pay bills?
2. Pick decent stocks for his clients?
3. Sell the Bentley?
Hi Sir,
ReplyDeleteI am sorry but I am not sure if I agree with you here. And there are several ways to look at it.
First, when investment funds compare its own performance to a benchmark or index, what they are really doing is trying to show potential investors what potential returns they will get from the fund in relation to the benchmark. In other words, the really important thing that is what returns the potential investor will get himself. Other things are mostly irrelevant.
And certainly, what restriction the constituents the benchmark has is not relevant to the investor. That is unless there is an imminent change in those restrictions. Since if the restrictions will stay the same, why should it matter then? Again, what he/she really cares is the returns the investor level collects.
Certainly, if the U.S. government really doesn't collect any taxes from corporation, the values of the companies in the index will go up. But that's just fictional and will not happen.
Remember, the tax discussion above centers around the constituents of the benchmarks being required to pay taxes, but not whether the investors will have to pay taxes on the return he/she collects. The latter is the real and relevant question to the investor.
And really, what isn't fair at all is to say the constituents of the index pay taxes, but that of the investment fund doesn't. This is simply not true. If you want to take the viewpoint of the constituents, then you should do the same with the Pabrai Funds. The companies that Pabrai Funds own stocks in also pay taxes. Essentially, Pabrai Funds is just an index, but one that is actively managed, of a number of tax paying corporations. That isn't much different from the S&P 500, Dow Jones or any other equity indexes in terms of whether the constituents pay taxes.
Please don't hesitate to correct me if I am missing anything or misunderstood you.
Dickson
Hi Dickson,
DeletePerhaps I can illustrate my point with an example.
Let's say that you and I are both managing money on behalf of some investors.
I have domiciled my investment entity in a tax haven and I pay no income tax. You on the other hand are domiciled in a country that pays income tax.
When I report my returns to my investors I report them on the basis of having paid no income tax (because I don't have to), when you report your results, they have to be reported after income tax. This situation gives me an enormous advantage over you because even if I perform exactly the same as you, my reported results will be better than yours by a multiple roughly approximating the formula I gave in the article above.