Not sure anyone noticed the back-up in rates during the past 2 weeks but this is very good for IVR's mark-to-market BV and also has a moderately positive impact on CPR speeds in the agency book, all else equal. Every 10 bps move in 5 yr swaps equals almost 20 cents per share in BV accretion. 5 yr swaps are now approaching 150 bps, up from a low of 107 bps just 1 month ago. The positive impact to BV will be seen in 4Q numbers if rates stay at these levels or move higher. Also, the improved swaps valuations means that IVR has to post less margin, which provides a greater liquidity buffer, should somewhat alleviate (largely unfounded) capital concerns, and frees up more resources to put to work buying assets. Some stabilization in credit markets should also ultimately help the mkt value of IVR's credit assets. While all of this is largely optical (as the mark to market losses on swaps will eventually go away along with the negative marks on the credit assets) it should nonetheless help stabilize the stock to some degree until it ultimately progresses towards its historical BV of around $19-20.
You "lost a lot of money" as in you sold your shares low or is it just an unrealized loss? If you haven't sold yet and you want out I suggest being patient and collecting the div double digit div waiting for a rally to sell in to. Unless you fear a dividend reduction there is no reason to sell. I think we'll be back above a 90 cent dividend within 6 months so if you haven't sold yet be patient and that way you won't lose money.
Thanks to all of you for that discussion on swaps. I have read that term for more than a year and now I finally have an inkling of what it really is and how it works. Very hard to understand the effective "kickback" to BV, but I even, finally, got that!
As to IVR it does seem undervalued, maybe; but it will probably be a long long time recovering and no room for more mistakes on hedging! They were forced to do the SO to dilute their over investment in swaps, instead of for a good reason, like exploiting higher interest rate opportunities.
I lost a lot of money on IVR but I also learned some very important investing lessons.
Exceptionally well explained. I'll add some further simplification:
IVR borrows at 30-day repo and buys long duration assets. This borrow short/lend long strategy locks in a pretty decent spread for IVR but bears substantial interest risks as higher short term interest rates would substantially erode this spread over time. Thus IVR buys hedges (floating-to-fixed interest rate swaps) to essentially match up the durations of its assets and liabilities. This results in a lower locked-in spread but lengthens the time period in which IVR earns this spread as it protects that spread from falling if short term rates rise. The cost of the swaps (and the resulting lower net interest margin) is already included in current earnings. However, in recent months the theoretical value of those hedges has fallen sharply (more so than the rise in the assets they are designed to hedge) as longer term interest rates have declined. This has no impact on the already locked-in spreads but it does impact GAAP BV via the Other Comprehensive Income account in Shareholder's Equity. Ultimately the negative mark to market hit gets amortized away as time goes by (as these hedges have a limited life span--currently averaging 4.5 yrs) and the result is gradual accretion to GAAP BV. Again, no impact to legacy net interest income as long as the assets remain on the books (as the spread has already been locked in). There are some negative consequences to IVR, however, in that as the hedge has depreciated in value the company has had to post additional margin with the dealers, which means a portion of equity capital is tied up and can't be used to buy spread-earning assets (which negatively impacts earnings). Lower non-agency asset values have had the same effect. This is likely already reflected in the current earnings run rate (assuming 80 cents is the new run rate, down from $1.00 per share--and I may be conservative in making that assumption).
Nevertheless, as interest rates rise the negative mark to market impact of the swaps decline reverses, which along with the passage of time leads to accretion in BV back to the historical BV of $19-20 per share. As that happens there is also likely some boost in earnings as more capital gets freed up to buy spread earning assets (but I view that as as gravy given the current $3.20 run rate represents a 22% dividend yield).
Hope that helps as well.
Thanks for the explanation. This kind of accounting is new to me, but it's definitely worth understanding the basics. Hopefully, I'm in the ballpark here:
The PV (present value) of the swaps is based on the time remaining. Assuming interest rates remain unchanged, the "hit" to BV is based on all future payments. As the months roll by, IVR makes the "fixed - float" payment and writes any losses as an expense (affecting their net interest earnings), but becasue the swaps have less time remaining, the PV of the swaps decline, and IVR is allowed to add that back to BV.
Simplifying a bit .... the MTM on the swaps is based on the present value (yield maintenance concept)of the fixed side based on current market rates. As market rates have gone down the higher fixed rate swaps have gone up in value thus causing losses in the swap portfolio - as the fixed swap rates are "worth" more than par.
The losses in the swaps are essentially being amortized each month via the expensing of the fixed rate interest payments (that are currently at higher than market rates). As each month goes by - all other things being equal - the PV of the yield maintenance declines because there is one less time period, and the book "loss" on the swaps is reduced. Current income reflects the actual fixed rates being paid; swap losses reflect the PV for the entire time period of the swap.
So the $4 figure previously cited (i haven't calculated that) accretes each time period throughout the four years and doesn't happen all at once at swap maturity.
As an example, if IVR has to pay someone $1MM a month in fixed rates versus an swapped floating rate of $800K that the counterparty has assumed - each month IVR expenses $200K more in interest expense than it would have without the swap. However, each month it does that reduces the portfolio swap loss by the same amount. Thus BV is ehanced by the "mark up" of the $200K for the period.
Thus it becomes apparent that swap hedges (not speculation) marked to market are an almost complete accounting fiction. Everything one needs to know (i.e. the effect on net interest margin of the hedges) flows though the income statement and the gyrations in the fair value of swaps is essentially nonsense.
I did simplify - hopefully it makes some sense.
I have followed this and agree with your analysis but my big worry is that they are going to get whipsawed because they did hedge this last round of post-offering purchases. But if the market is following the second half of 2010 playbook (risk on in other words) I believe a book of $18 and pps of $17 are possible. There are so many trapped longs (me included) however in the 19s and 18s that IVR will never see $18 again IMHO.
Historical BV is $19-20. That's where GAAP BV is ultimately going. The stock should also ultimately head in that direction plus pay large dividends along the way. Historical BV was close to $21 before the last SPO at around $18+ so the $19-20 is the current stated BV.
According to the JMP conference, the proceeds from the August offering were used it to buy agency MBSs, and they did not add any hedges; they admitted that they were already over-hedged.
As for being a trapped long: I too had a cost basis at $21; I added more at under $14. I can always sell the "extra" shares later (e.g. Dec). Their 80 cent dividend looks okay, and their BV should recover more between now and then. In addition, I can always sell before x-div and buy back on or after x-div and use the "wash rule" to lower my cost basis. It may take another quarter or two, but I won't be underwater much longer.
The $19 to $20 BV may be a long time coming due to the low priced SPO. Especially since it increased the dividend burden. You are correct about the direction of BV though, TNX may hit 30 again my end of 4q just like it did last year. It's freaky how well the two year's TNX charts overlay each other thus far.