The tax simplicity of tax deferred accounts

A common question is whether to put money in a 401K or IRA or not. Unfortunately, there are a daunting number of options and an even more bewildering battery of rules for each choice.

However, one often ignored but compelling reason to put money in a tax deferred account (401k, 403b, IRA) is simplicity. Tax simplicity in particular.

If you're just starting to invest or your accounts are relatively simple, then this may seem like a moot point.
But as time goes on and your portfolio grows and becomes more diverse, you will soon find that filing your taxes and keeping all the bookwork starts to take a life of its own. A time consuming, minutae filled life.
And typically it only gets worse over time.

Assume you are a long term investor and buy a mutual fund and just hold it for 7 years. And you reinvest all dividends and capital gains. What can be so bad? It's the regular dividend and capital gains distributions.

Let's take the Vanguard S&P 500 Index VFINX, a widely held mutual fund (the third largest as of Mar 2007). It is a well-behaved fund as it has few if any capital gains distributrions, from selling appreciated stock, in part because it is a growing fund (so all they do is buy more) and in part because it is very well run. But it does have quarterly dividend payouts. (As an aside, the Vanguard website on VFINX indicates the unrealized capital gain in the fund price is very significant at 38% of the NAV, meaning they are holding lots of appreciated stock).

Let's outline the problem(s) of the dividend payout.
- On every dividend payout, you have to add them all up and pay taxes at the end of the year. No big deal.
- Each time you reinvested dividends, you buy a bit more of the fund, at the price on the dividend payout. Not a big the first time, but it starts to get worse.

After 5 years, you have made 21 different purchases, 20 of which were dividend reinvestments for small amounts at different prices and different amounts. You essentially have 21 positions in VFINX. Yow.

Eventually, if you sell all of your holdings, at tax time it is a tedious calculation mess, because you are essentially selling out of 21 different positions, all of VFINX. But at least you are done with that investment once and for all, from a tax standpoint.

But if you only sell a fraction of your holdings, say to rebalance, then you have a mess now and bigger mess later. You have to decide from a tax standpoint which shares you sold - (a) the original shares, or (b) the most recent shares bought by divident reinvestment or (c)perhaps you sold all shares proportionally. The simplest taxwise is (a) but this is likely the worse from a tax payment standpoint as the original shares have appreciated the most. If the price has gone up somewhat steadily you might choose (b) to minimize your tax burden, but might want to sell shares you've had for at least 12 months for LT gains. Finally, I cannot recommend (c) under any circumstance as it complicates the books now and especially in the future.

Now imaging if you buy or sell more shares every year moving forward due to rebalancing concerns. As you can see the bookkeeping just keeps getting uglier. After 4 years of rebalancing, it isn't hard to imagine having 30+ positions of VFINX, many of which had already had some buying and selling. It takes several years to reach that point, but if all goes well for you, you will reach that point.

The other alternative is to not reinvest dividends or cap gains. However, this creates small bits of cash in your portfolio at regular intervals, which is hard to invest. If you want to cash for cash flow, that's OK, but it is irregular in that you can't predict cap gains very well. And if you are constantly pulling out cash, it is harder to determine return. So I assume you want to reinvest, if at all possible.

In contrast to all this, having a tax deferred account is very simple.

1. If is a Roth, you never pay taxes. This is the simplest. The Roth is my favorite vehicle, though both the Roth and the pretax 401k/IRA have roughly equal tradeoffs.
   
2. If it a pretax IRA or pretax 401k, you pay ordinary income on everything you withdraw. Again very simple from a tax stand point.
   
3. If it is a post tax IRA or 401k, it is a bit more compliated but again not too bad. And you get to defer all this work until you actually withdraw, so it is still a procrastinators dream.
   

Catching the falling knife

Today (Feb 27, 2007) saw a 10% drop in the China stock market, which caused at least a 3% drop in most other major stock markets. On a serious decline or plunge, the experts warn us not to invest as the various markets plunged, saying instead to wait for this correction to stablize. The common phrase I've heard (over the last decade no less) is that investing in a plunging market is "like trying to catch a falling knife."

Yet, I've done quite well catching that knife.

So today, I saw the market was down in the morning. Then I happened to recheck the market and realized how far the market had dropped by 3 PM EST (noon PST) and decided to do some knife catching. I picked up various ETFs at near the lows of the day (FXI, ILF, VTI, RWX). It remains to be seen if the market rebounds, and whether I got bargains or the market continues to fall and I got cut by the knife.

I also need to study how often the knife falls below it's initial plunge. I got the impression that todays' fall was simply a sell off with no particularly compelling catalyst. Hence my hope that markets recover.

Update the next day 2/28:

Pretty much everything was up today, the next day. China recovered 4% and the international markets were up about a 1%. The US market was up a tad, too.

In the SJ Mercury lead article, they mentioned there had been 14 single day declines worse than yesterdays decline of 3.3% in the last 40 years. And in 9 of the 14 cases, the markets were up the next day. So you can catch the knife more often than not.

Asset allocation in the presence of unrealized gains

This is an oft ignored question: should your asset allocation be affected by the amounts of the unrealized capital gains in your position. Let me give an example. Assume for simplicity you want 70% stocks and 30% cash.

Case 1: you have $10,000 of cash to invest. You pretty clearly put 7K in stocks and 3K in cash.

Case 2: you invested in stocks many years ago and they have done well. You have large unrealized, capital gains in your stocks and they are now worth 10X their original investment, namely $7K. If you were to sell all your stocks, you would be taxed and would be left with a siginificantly smaller sum. Or you can just sell enough stock to bring your allocation in line with the 70-30 allocation.

A coworker had asked this a few years back and I came to the pleasantly simple conclusion:

No, you should safely ignore all this and allocate oblivious to unrealized gains.

The short summary (which I need to elaborate on later) is that by considering the incremental return-risk tradeoff of a given allocation, the behaviour of your investments is blind to how much unrealized capital gain there is. Thus you invest blind to this. Thank goodness.

The cost of borrowing

The opportunity cost of borrowing

I'm a saver at heart. A dollar saved is another dollar I can mismanage investing. Yet, given the massive amount of credit card debt and home refinance debt of this country, there's a lot of non-savers out there. I have to believe that those with a spend-now attitude, the so called "gold collar" workers, simply don't know how costly it is to spend beyond your means. Yet sadly, I think even if they did, they wouldn't care.

I think one of the problems is non-symmetry of a dollar owed versus a dollar gained. Consider the following two scenarios.

1) Happy: If I told you I would give you an extra $1000 a month, how happy would you be?

2) Sad: In contrast, if I told you I'm going to increase your bills/housing cost by an extra $1000 a month, how unhappy would you be?

I'm betting you'd be a lot unhappier than happier. Yet if you borrow money (aka use a credit card and don't pay it off) you get your happy day now but suffer the (much) sadder days later. And there's more sad days than happy days. Doesn't seem like such a good deal does it?

But since this a column on investing, let's do the math.

Things are a bit different depending on whether you have the money or not. For the sake of simplicity let's assume you can either buy a $18000 econo car or spend another $10000 for a new $28000 nicer car. Clearly you'll be at least $10K poorer if you buy the nicer car. The question boils down to, how much does spending the extra $10,000 cost you down the line.

Case A: you have to borrow the extra $10K. (How you get the first $18K doesn't matter since in both cases it is treated the same).

[Nicer car] you have borrow the $10K and get a 8% loan for 5 years. This loan is short enough that we'll ignore the compounding of the interest. The total interest you'll pay is 10K * 8%/year * 5 years * 1/2 = $2K in interest. (The 1/2 comes in because after you're steadily paying down the principal and over the loan lifetime [5 years], on average you only have 1/2 the original principal building up interest. In reality, the 1/2 term would be slighly higher, say 0.55 or 0.60 since 8% * 5 yrs = 40% isn't completely negligible compare to the original loan). In general, for short term low interest loans, the interest you pay is Principal * Interest rate * Loan Duration * 1/2.

[Econo car] Instead, if you bought the econo-car and had invested the money you used for the loan payments, you would have gotten investment gains. Let's say you could earn a paltry 6% on your money. Over 5 years, again ignoring compounding the investment return you would get is $10K * 6%/year * 5 years * 1/2 = $1500. (Again the 1/2 is because initially you have no money invested and it linearly builds up, so on average over the 5 years you have 1/2 the money invested earning returns).

So with our numbers, that buying the more expensive car means you'll be $13,500 poorer after 5 years. If you don't like the loan rate and investment return I used, the formula is

Principal * (Loan Interest + Investment Return) * Loan Duration * 1/2.

In short it's a double whammy when you borrow, because instead of having the money grow, you're not only unable to invest, you're losing some of it to pay off the loan interest. And you never get this money back.

To beat this over your head, let's say instead you're borrowing with a credit card, which charges 18%. Then, borrowing this $10K (assuming you pay it off after 5 years, which is iffy) would cost you another $1K in lost money or a total of $4500. In reality, by ignoring compounding the numbers come out friendlier. So let's say that you'll really be down an extra $6K in addition to the $10K. Namely, not only are you $10K poorer, you have additionally agreed to dispose of out an extra $100 a month, every month, just to buy the nicer car. How much nicer was that car again?

Case B: you have the money.

Here, it's simpler. If you buy the cheaper car, you can invest that money and earn your 6% (or 8% or 10%) over 5 years, so after 5 years you are $14K to $16K richer. Here, I'm assuming compounding since the money is all there from the beginning.

So in this case the cost is
Principal * (Investment Return) * Loan Duration .

The last whammy

The other whammy is that if you spend more for a car or TV or dinner or most anything "fun", at the end of 5 years, that item usually isn't worth any more than if you just saved the money.
So in most cases, you really are out the $10K + the non-neglible cost of borrowing.

A few more notes

When does the of the interest rate matter? E.g. for a 2 year 6% loan, compounding is neglible but for a 30 year 7% mortgage, most of what you pay is the

Another case which a pesky reader asked about is what if you borrow when you don't have to and invest the money in something with more return than the loan interest rate. Again assuming a lightweight loan that you continually pay down, then your return on paper (aka you are making money, here) is:
Principal * (Investment return - interest rate) * Duration * 1/2.

But if you factor in taxes, things don't look so nice. You always pay taxes on investment gains, but for almost all loans (except mortgages) you get no tax break. Let your overall tax rate be TaxRate, so you get to keep (1-TaxRate) of your investments assuming you sell your investments to pay back the loan.

Principal * [(Investment return * (1 - TaxRate)) - interest rate] * Duration * 1/2.

If the loan was a balloon payment in which you pay back everything at the end, then the 1/2 goes away. This formula also makes intuitive sense.

International real estate funds/REITS/ETFs

The asset class I'm most excited about recently is foreign real estate. I've been waiting for a mutual fund or ETF to invest in a variety of international real-estate instruments (they are not called REITS in other country, but they are the equivalent to REITS). Let me call these investments: "Intl-REIT"s versus "US-REIT"s, even though in all cases you are investing a basket of REIT-like securities.

I'm a big believer in low cost index funds or ETFs and had been waiting for such a Intl-REIT ETF.
And in late Dec 2006, the first low cost ETF based on Intl-REIT opened up: RWX which tries to mirror the Dow Jones Wilshire International Real Estate index. More information on RWX is at this MarketWatch article. http://etf.seekingalpha.com/article/23080 indicates that roughly 18% of the assets are in each of Australia, the UK and Japan, with other well developed (not emerging market) regions filling out the top 10 countries. I first bought this ETF in Jan and will likely add more to my holdings.
And while it has done freakishly well in the first 4 weeks of 2007, so far, I'm don't put much weight in recent short term performance.

However, there are several Intl-REIT mutual funds that have been around for many years. I probably should have looked harder last year to find these. Anyways, here are some links on other intl real estate mutual funds/ETFs.

1) The Kiplinger's article
www.kiplinger.com/columns/balance/archive/2006/balance1226.html
is a good starting point, as it mentions that Intl-REIT correlates with other investment, including US-stocks, better than you (and I) might have expected. It also mentions several worth while funds. I really like the Cohen & Steers Asia Pacific Real Estate Securities (APFAX), but I'm put off by their 4.5% initial fee. Maybe I can get my financial advisor to lower this.

2) This Smart Money article discusses the three well known Intl-REIT funds. I personally just bought some of the Alpine fund today, largely due to its low expenses and long track record. Also from my brokerage (Schwab), it was no-load (no transaction charge or fee).

3) There are apparently more international real estate ETFs in the works