Monday, August 30, 2010


In my opinion, the risk/reward for fixed income vs. stocks is heavily weighted toward stocks.

So you want to invest in long-dated bonds? Here are a few things worth mulling over.

With a yield of about 3.60%, investors are talking on far more risk than they might imagine. Without going into the theoretical construct of duration and convexity, let's try to put this risk in simple terms.

If in one year, the 30-year yield declines another 50 basis points, then your bond would have appreciated 9.52% in price. Added to the 3.60% in coupon, you earned 13.12% in that year.

That is the reason why TBT has been getting massacred. Rate drops are magnified 2 for 1, and the inverse leveraged ETF continues to drop rapidly.

However, if the rate backs up (yield increases) 50 basis points, then your bond value will decline 8.44%. Offset that by the 3.6% dividend and your total loss is 4.84%.

Let's take it a step further and say that the 30-year backs up 1% in the next two years. Your bond value will decline 15.66% offset by the 7.2% of dividends that you earned. Of course, it does get worse the larger the rate increase in a shorter amount of time.

At best, if you hold that bond to maturity, you will earn 108%. That does not assume reinvesting of coupons and does assume that you hold it to maturity. Most bond holders won't be doing so. Most retirees will be dead by then.

Finally, there is good research from Brian Reynolds at WJB Capital, who is one of the finest fixed-income analysts with few peers. He is saying that the credit bull market is continuing and that equity investors do not believe it, and hence we are in a stock market correction within a bull market. Eventually, his research indicates that the credit market wins out and stocks will rise.

What most people don't quite understand is the risk that they are taking by buying long-dated debt at these record low yields.

I expect the M&A calendar to become more crowded as stock valuations remain cheap and cash plentiful at low rates.

No comments: