Sunday, October 25, 2009

On AGNC's Dividend

An article recently suggested that AGNC's dividends were doomed because they were supported in part with capital gains. Since my own exposure to AGNC is modest, I'm not particularly keen to debate the merits of the prediction that AGNC's dividend might fall. A reader asked what The Jaded Consumer had to say on AGNC's dividend, so here it is: I don't know.

AGNC is capable of working with 10x leverage without violating its governing documents. Like a bank, AGNC makes money on interest spreads (the interest earned less the cost of borrowing, as multiplied by leverage). To ask whether AGNC can keep up its dividend is to ask whether AGNC can borrow more cheaply than it lends to such an extent that income stays up. AGNC has demonstrated creativity, though: when volatility went through the roof last year, AGNC shed leverage for protection and lost the ability to make money from multiplied interest spreads (at the end of 2008, leverage was 5.2x) -- but earned money from the volatility due to options activity (in effect, covered calls). Creative approaches to financial management in the face of changing conditions has enabled AGNC to earn money at a much better rate than onlookers might have guessed on the basis of interest rate spreads alone.

I don't dare predict AGNC's dividends -- any more than I'd try to predict specific interest rate spreads or prudent leverage levels -- but I offer the quality of AGNC's management (rented from ACAS), as evidenced over AGNC's history, as a counter-argument to the thesis that some particular indicator (like the appearance of capital gain in AGNC's income) proves that AGNC's dividend is doomed. Sure, capital gain in fixed instruments may be cyclical and therefore doomed to a finite life, but this doesn't mean AGNC can't make money in other ways. With government stepping in to guarantee the obligations of Fannie and Freddie (for example), increasing leverage beyond last year's levels may be more sensible.

Let's have a look at ACAS' management of AGNC, though: ACAS, like many other shareholders, have very much wanted high current dividends during the liquidity crisis of the last year. ACAS has benefited from AGNC-created liquidity, and if the train slows down for cyclical reasons it's of little harm to ACAS. My real question is whether AGNC can raise more capital at its current above-IPO prices. ACAS benefits from capital increase because its management fee is based on the size of funds under management. However, ACAS seems to make quite a bit more at the moment as a shareholder than as a manager.

(To those who worry about AGNC holding ARM-based assets: imagine AGNC holding fixed-mortgage assets in a climate of increasing interest rates, and ask whether holding ARMs is risky or is a hedge. Frankly, I believe home mortgage interest rates are already increasing and that having begun moving from fixed-rate holdings early in the year is a Good Thing™.)

The fact that ACAS has been able to generate outsized returns from AGNC could be dismissed as some kind of fluke, but I doubt success entirely an accident. ACAS may not be able to replicate indefinitely the crazy returns it's produced to date from AGNC, but this doesn't mean that market-beating returns are dead. The fact that AGNC's investment mandate has such rigid restrictions -- it's a REIT, for example, and invests in agency-backed mortgages (and obligations comprised of such mortgages) -- means that it will be subject to a host of limitations peculiar to those investments, constraining what ACAS can do to wring profit from it. However, creativity demonstrated thus far suggests ACAS can do quite a bit with AGNC under various conditions to create value. The only question is whether the value ACAS can wring from AGNC has peaked. Since the performance of agency backed mortgage investments isn't stable over time I'd expect ACAS to derive an unstable benefit from AGNC even if it continued to produce market-beating returns.

So, has the dividend peaked? Who knows. The question is whether market-beating returns have ceased, and the answer there is maybe not. And let's assume for a moment that the day comes that an investor becomes convinced the dividend must be decreased. Is this a reason to bail from AGNC? Only if you have someplace even more market-beating to put your money.


IandW said...

Thanks for your analysis! Even if they eventually come down to half the dividend yield, 9-10% would still pretty juicy, above what most other REITs yield. Of course, AGNC's share price would take a beating. I've been down this road before (FRO - Frontline Ltd).

Anonymous said...

You will make in one year , what you would have made in 4 years at 4%, that puts it in perspective lads.