JustCurious said:

Nords, you excluded (1) closing costs, (2) trading costs, (3) taxes. So your results are flawed. Not to mention that "past performance does not guarantee future results."

Ye gods, man, are you sure that you read the whole thread? Here, wait, I'll get the relevant parts of it for you:

Nords said:

Dividends (~1%) & cap gains (if any) will be reinvested free of charge. The closing price on 8 October 2004 was $110.06.

Since closing costs will be recouped by the lower payments in just eight months, I'm not going to count that against the returns. The Fidelity brokerage fee is minimal (one really big trade in a frequently-traded account) so I'm not counting that either.

Nords said:

We refinanced again this month-- barely nine months after the last refinance (which has already paid for itself). This time we went from NFCU's 5.5% to 5.375%. Zero points, less than 80% LTV, no cash out, applying online, no documentation or appraisal, and haggling with the title company brought the total refinancing cost down to $1151. Our payments dropped by $44 so this time it'll take 27 months to pay back the refinancing costs.

Back in Oct 2004 (the beginning of this thread) we moved a large chunk of our portfolio, including the refinancing proceeds, into the S&P600/Barra Small-cap Value ETF (IJS). I'm tracking this on a spreadsheet with the assumption that we subtract 15% from dividends (to pay taxes) and reinvest the remainder (which Fidelity brokerage does for free). The ETF is up 18% since then and reinvesting dividends has boosted the return to 18.7%. Sure it looks smart now, but let's give this another 29 years before making a judgment call. We could look equally stupid next spring but we'll keep reinvesting those after-tax dividends.

Nords said:

Annual update: The latest quarterly dividend was just paid and we're up 31% after-tax over two years or about 14.5% APY. I'm assuming that dividends are taxed at the 15% rate and reinvested.

Nords said:

At

Drip Guy's question, here's an update. Three months is meaningless in the context of a 30-year mortgage but this might really really be the top.

IJS paid out a little over 24 cents/share last week and we're now up a total of 40% over 27 months. That looks like about 16% APY even after paying taxes at the 15% rate and reinvesting, although I'm not sure how much of that payout was cap gains.

And I'll recap:

JustCurious said:

Nords, you excluded (1) closing costs

The lower mortgage payments have more than taken care of the closing costs. I'll also point out that the total sum of all the closing costs we've paid (getting the loan and then refinancing it) are less than 1% of the loan a couple percent of the profits. Hawaii mortgages are pretty hefty and we haggled hard on the closing costs.

JustCurious said:

So far that's a Fidelity commission of three $8 trades. We could've done it one trade but I suspect that the share volume would've driven up the price. Since we paid that commission (Oct 2004) all dividends have been reinvested for free.

JustCurious said:

Well, so far I've been paying the taxes out of other funds, as I suspect that anyone else would do in the real world. I'm not going to liquidate shares for an $8 commission to pay a few hundred in taxes. But I've tracked the numbers on a spreadsheet, and on that spreadsheet I've been paying tax on the dividends at what I feel is an extremely conservative 15% rate before reinvesting them. The returns I've quoted in this thread are after-tax returns. If dividend tax rates change then I'll change the spreadsheet.

JustCurious said:

So your results are flawed.

Yes, and extremely profitable. I suspect that the "flaws" don't invalidate the two points that have been made so far:

- mortgage arbitrage can be profitable, and

- a bigger ER portfolio is more survivable.

Of course we'll know the full results in another 28 years.

JustCurious said:

Not to mention that "past performance does not guarantee future results."

You're absolutely right. The best data I could find was Dimson & Marsh's study on a century of investment returns claiming that there's a small-cap value premium. I feel that conclusion has been adequately validated by Bernstein and others. Of course the next couple decades could prove that past is no longer prologue, but I'm going with the best information available.

When you have a better system then I'd be happy to use it.