In reply to Peter's comment, I offer:
Thanks for reading, and for the kind words. Unfortunately, I started buying ACAS north of 40, so even though I (approximately) doubled at $1.80, I'm still not in the black yet on ACAS.
As to ACAS' sensitivity to the impact of 150 basis points' move on the performance of AGNC, it looks like this has two (or three) parts. First, AGNC uses some hedging techniques and I'm not sure what their impact will be. Back in 2008, when liquidity was tight, AGNC's hedges made more money than its investments. The other side of the AGNC performance is that AGNC's profits aren't based on rates, but rate spreads. AGNC finances its portfolio with repurchase agreements. Since its portfolio is liquid – that is, the portfolio is full of government-backed investments for which there is a ready market and pretty clear valuation – ACAS has (as AGNC's manager) been able to leverage the portfolio, and change the debt:equity ratio as it pleased based on the environment. So, the first part of the equation – the impact of a 150-basis-point increase on AGNC's profitability – is something that is both non-trivial and outside my realm of knowledge. What are the odds of a rate-spread inversion? I'm not really the right person to ask. The last (either second or third, depending how you are counting) part of the question asks: what is the impact of AGNC's fortunes on ACAS. Now that ACAS isn't a big shareholder of AGNC (it's sold its stake, largely at a gain; the latest 10-Q shows no unrealized AGNC-related gains and no AGNC holdings), the real impact is in its monthly management fee. This management fee is based on AGNC's assets under management. Based on the math here (more than half a penny per share per quarter added to ACAS' management fees due to the described issuance) and here (another issuance adds $0.07 per share per quarter to ACAS' revenues), it seems ACAS' revenues from AGNC are likely more sensitive to issuances than to revenues. The three months ended June 30 showed AGNC paying over $12.4m in management fees to ACAS, which based on the nearly 354m shares outstanding at the end of the last quarter reported amounts to over 3.5¢ per share (annualizing to over 14¢ per share). Profits of REITs must be distributed, so they don't generally boost assets much – and ACAS' management fee is based on assets under management, not profits. AGNC must avoid losses to protect its assets under management, but if management is as nimble in reacting to changes in the macro-environment as it has been in the past, the probability is against being stuck for a long time in a money-losing environment.
The dividend historically was paid based on BDC requirements. ACAS deliberately failed a RIC test so that its loss carryforward would not be lost (BDCs can't carry forward all the kinds of losses that C-corps can). ACAS at present has no dividend requirement. By the time ACAS has a taxable profit, I expect it'll arrange to pass RIC tests again. The question is, when will that be? Since the answer depends in part on when ACAS realizes gains and when it realizes losses, it's possible that management controls this much more than would be predicted based on projections based on the macro-environment. Essentially, I wonder what management's preference is, and whether management makes a priority out of any particular tax status. Without that knowledge, I don't think dividend timing projections will mean much.
I think the best place to mine for insight into the future of dividends may be the statements of management on quarterly calls. Still, developments since the time of the calls are likely to impact management's current thinking.
As the market is hit, ACAS' portfolio's comparables will be hit, and with it theoretically ACAS' NAV. This is an interesting time: ACAS announces another great quarter of NAV increase, and its share value is punished. Does the punishment fit the market, or is it excessive? We can't know until next quarter's announcements. Will NAV still move in the right direction? NOI?
ACAS will continue to be volatile. I still view its below-NAV pricing as a value opportunity. I think ACAS' solvency makes it appropriate to consider as a long-term investment, and long-term investments in things so diverse as to represent the broader economy are theoretically a bet on the broader economy itself. The question is, when is it appropriate to make big bets on the broader economy? Buffett has said that's what he's done, so his view is now public. For those of us with a shorter time frame, might we be better able to time entry? I certainly have very limited success timing buys and sells.
I'm much better at spotting long-term opportunity: Apple in 1998, for example. For most of the time from then on most of my assets were in Apple -- my best bet. But as volatile as that's been, there's as much opportunity to have lost money with bad timing as in ACAS. The reason I own much less now than I did was that I got bored with price stagnation during a plateau and lost shares to the exercise of covered calls when the shares started moving again. Heightening exposure to volatility risk might not be the best strategy for the long-term investor. Assuming the underlying thesis is good, the only safety may be adopting a Buffett-like investment horizon.