"If You Think You Can Beat Buy and Hold, Try Looking at Taxes"

Standard and Poor's gives SP500 returns with and without dividends back to 1990. Here are the returns at year ends, without and with dividends:

date       1yr rt  total rt
12/1990 -0.066 -0.031
12/1991  0.263  0.305
12/1992  0.045  0.076
12/1993  0.071  0.101
12/1994 -0.015  0.013
12/1995  0.341  0.376
12/1996  0.203  0.230
12/1997  0.310  0.334
12/1998  0.267  0.286
12/1999  0.195  0.210
12/2000 -0.101 -0.091
12/2001 -0.130 -0.119
12/2002 -0.234 -0.221
12/2003  0.264  0.287
12/2004  0.090  0.109
12/2005  0.030  0.049
12/2006  0.136  0.158

Mr. B. Holder (BH) puts $100,000 into the SP500 index 1/1/90, reinvests dividends into the index at year end, and pays tax on dividends out of his salary as school janitor. BH sweeps into retardment 12/2006, and notices that without taxes his initial investment became $555,637, including reinvestment of $22,271 in dividends along the way.

For this exercise, let's assume that long-term capital gains and dividends are both taxed 20% (they are currently taxed less, have been taxed more in the past, and will no doubt be more in the future). When BH cashes out 12/06, he gets to contribute $91,128 and keeps

Also in 1990, recent refugee Natasha Otradeskaya (NO) plows $100,000 of hard stolen capital into a U.S. trading account. Like others from her country, NO has degrees in math, physics, and nuclear medicine, but unlike competitors in the US and A, does well enough with trading (after slippage, commissions, and opportunity costs) to exactly equal yearly SP500 returns including dividends. Exhausted after 16 years and 10^16 quantitative studies, she decides to look back at her results:

date      100000     P/L     tax        net
12/1990   96896   -3104       0
12/1991 126416  29520   7925  118491
12/1992 127526    1110    333  127193
12/1993 140004  12478   3743  136261
12/1994 138060   -1945        0  138060
12/1995 189939  51880 14980 174959
12/1996 215129  25190   7557 207572
12/1997 276825  61696 18509 258316
12/1998 332140  55315 16594 315546
12/1999 381941  49801 14940 367001
12/2000 333588 -48354        0 333588
12/2001 293938 -39650        0 293938
12/2002 228973 -64965        0 228973
12/2003 294665  65692        0 294665
12/2004 326724  32060        0 326724
12/2005 342770  16045        0 342770
12/2006 396907  54137   4490 392417 <<<<Final balance

NO trades full-time 100% of her account every year, and pays income tax of 30% on her gains (the calculation carries forward losses, which are balanced against future gains; which is the case for a trader who has no other income tax to offset capital losses. Note also that losses carried forward from 00-02 are not used up until 2006). Thus she is unable to reinvest what is lost to taxes, and these amounts do not compound.

In the end, BH cleans up: He gets paid an extra $72,093 for not doing studies or trading frequently, and uses it to order a new bride who happens to be NO's grand-niece from the old country.

J T Holley adds: 

The old gray mare ain't what she used to be. You can actually get no load low M&E annuities that aren't as bad as you would think. Most people think though that investing should be free and that one bp is too much.

Vanguard's isn't in the example above but its M&E is .20 basis points with a .10 basis point administration charge and the S&P 500 clone sub account is .44 expense, bringing the total to 74 beeps. Anyone counting this out remember that annuities are like mutual funds and take fees out daily on an annual basis.

I'll take the ordinary income later in life over the stg/ltg now and avoid Uncle Sam. Even with the 74 beep vig annually that is better than the taxable stg/ltg combination paying taxes now, that the S&P 500 tax toll can be annual. I like the compounding and pay later acting as a quasi spigot trust if you have enough time and persistence.

Steve is right though. There are tons of annuities that have a lot higher costs. The industry average is around 250 beeps in M&E and charges, but even these higher vigged charges show the tax deferral outperforming the pay as you go in a taxable account if you hold long enough to let it happen.

Also, annuities have a 10% penalty much like IRA's if you yank out before 59 ½, unless you 72T. I would recommend neither unless "liquidation" is really needed.

Mr. Sears might help us here. I'll defer all annuity and running knowledge to him.

Steve Leslie writes:

ETFs are not fairly new. Diamonds and Spiders have been around for a decade or more. And irrespective of their longevity if one's goal is to replicate an event using an ETF, what difference does it make how long they have been around? This is not Morningstar; as we know track records do not matter.

There is a big misunderstanding as to the elimination of estate tax law in 2010 and the questions surrounding it. The reason there is a gap in 2010 is that it is my understanding tax law cannot be written for more than 10 years. So it is clear that the next administration will rewrite the tax code. Who knows what they will come up with then?

There are some very low cost variable annuities which stripped down to the bare minimum may have an attractiveness. When I was a broker (sounds like my father) the lowest cost one we sold was the Nationwide Best of America. Still, be very careful that there are some pretty good back end charges if you redeem the annuity. I think there are some no load annuities with low M&E perhaps in the Vanguard family. M&E is mortality and expenses.

This is where all the hidden charges show up. You pay for that death benefit. And the charges are rarely below 1% and then you add management fees of the funds on top of that. The management fees can be higher in annuities if they are funds outside the annuities family of funds. You pay to play. And once again there is no stepped up cost basis on date of death. That is an extremely expensive thing to have the heirs have to face should that happen.

I still argue there is no perfect solution, however, ETFs stack up very well against no load index funds, in performance for expenses and after tax return ease of ownership and liquidity. One added thought; they are a heck of a lot easier to track the cost basis for reinvested dividends than mutual funds. You do not get average cost basis. First in first out, etc.

I am not sure about the dividends being left in cash as Mr. Sasmor suggests. I thought by definition they try to be as fully invested as possible. Not even open-end index mutual funds are fully invested at all times. They do by the mechanics of the machine have to keep some cash on hand. And in open ended mutual funds, when a run on liquidity happens they must sell stocks accordingly to handle redemptions. Whereas with closed end funds this does not happen. There may be times when closed end funds trade at a discount to the NAV. Open ended mutuals are marked to the market on the end of business. And purchases are restricted to last day's closing price.

Bill Rafter adds: 

Regularly I get an idea from an investment professional who typically does not do any quantitative research, but has a "hunch" that he thinks will work. More often than not the investment pro got the idea from some huckster. Investment professionals are really no different (nor less vulnerable) than the public at large. To keep the investment professional devoted to quality research, I have to disprove the cockamamie hunches, lest he wander off following outlandish ideas.

The latest is the "gravity idea" based on "celestial mechanics". Every good con job contains a certain amount of truth, and this one evokes Isaac Newton, Kepler, and some other luminaries. The celestial mechanics is BS to make the huckster sound important and establish him in an intellectual pecking order. I think this is something they teach in huckster school on the premise that the mark has to look up to the huckster.

The system the huckster is peddling is really just based on tides, which of course is a function of celestial mechanics. But neither the huckster nor the mark knows that. I can tell the huckster doesn't know that because he claims to use both the celestial cycles and the tides, which in his case is redundant.

The huckster uses tide schedules from a point in Delaware Bay. I cannot really find out why because all conversation has to go through the mark. Apparently the mark has told the huckster that he has had conversations with another one of his confidants (me) who is skeptical. This gets the huckster's back up: how dare the mark give any credibility to non-believers! Another lesson from huckster school: get indignant when challenged and threaten to cut off communications. The mark always wants what he cannot have.

The huckster really has the mark convinced. I try all sorts of arguments, like why we cannot find any evidence of these cycles using Fourier analysis. And if the idea were true, why isn't there more volatility during the "spring tides" when the earth is at perigee? However, I'm getting nowhere because the huckster has a track record with great performance. Of course there are reasons why he cannot show it, so no one has seen it.

Well I don't want to go searching for the Delaware Bay tide schedule. And since I live near the seashore, I have the Barnegat Light tide tables bookmarked on the PC. That data are neatly presented in a spreadsheet with the date and four columns representing the two daily high and low tides. The similarity with bar-chart data is overwhelming, so I copy the data, import it into my charting software and have it displayed as open, high, low and close. I send that chart back to the investment professional with my commentary that the highs on the 2nd of the month will be higher than the highs on the 16th, and the lows of the 23rd will be lower than the lows of the 9th.

At the end of the month the investment professional (i.e. the mark) gives me a call and tells me that my trading signals were right on target, and the gravity man (i.e. the huckster) complements me on having mastered the technique in such a short period of time. I know the trading signals did not work and I suspect that the huckster is now using me to validate his own importance. This must be another lesson from huckster class: if confronted with an outside challenge, turn the challenge around to support your own claims.

I have no choice but to have a personal meeting with the investment professional and show him exactly what I did, which takes about two minutes. He really doesn't want to believe me, but when he sees that my tidal signals did not work and then realizes that the gravity man is just playing him, I finally win him over.

Hey, anyone want to buy my gravity system?

Steve B. adds:

The con game is not so much about the con be it in financial services, health and beauty, or the red-hot real estate (mortgage) market. The con game is about the mark - the one who must get conned for the game to work. The con in this game has the advantage of instant feedback from the mark. The mark is spotted by the type of work he does, the kind of car he drives, his or her age. These are just general and to get specifics he needs to talk directly with the mark.

How is it that the con can come to know about us? We provide him the information via physics and other shamsters. But in this case we feel we are dealing with a respectable person and our guards are down. Basically, one starts with general statements and through feedback (verbal and body) the con figures out which of his general statements are specific to you.

We all have the desire to trust our fellow human and the con operates with this advantage. You may not always know when the con is on but if you feel the flattery (and a feeling you may miss the boat) then you may be a mark.

From Henry Gifford:

While it seems so many con games include some true physics in the story, the physics are usually redundant (as already mentioned) or not relevant. My favorite is energy-saving devices that incorporate the use of "bipolar" magnets. As all magnets have two poles, the statement is meaningless yet true. 





Speak your mind


Resources & Links