The Curious Case of DWS Investments

Includes: DB, SRO, SRQ
by: Gwailo

Investment managers at Deutsche Bank's DWS mutual fund complex are trying desperately to bury their mistakes. And no wonder. Under their guidance, the DWS RREEF Real Estate Fund II (closed-end fund ticker "SRO") lost 91.6% of its value in the fourth quarter of 2008, incinerating about $340 million of investor equity, while its smaller twin fund I ("SRQ"), down only 82%, burned through another $200 million. These losses may have set some new record for the rapid destruction of shareholder value by professional asset managers, and they earned both funds a solid 1-star ("worst") rating.

Now both funds have asked their shareholders to approve plans for liquidation: the remaining assets are to be sold, the proceeds distributed, and the funds will cease to exist. SRO and SRQ will only remain as fading, painful memories for their unhappy investors. [Disclosure: The author is a retired accounting professor who bought 500 shares of SRQ for an IRA account back in July 2004. The shares are now worth about 1/10th of what they had cost.]

But an unlikely hero has come to the rescue. The liquidation plans emerged from a Board of Directors meeting on March 11, shortly after SEC filings by one Stewart Horejsi had announced the purchase of about 7% of SRQ and 5% of SRO. Horejsi, a wealthy activist investor who has wrested control over several other closed-end funds in recent years, proposed to terminate DWS/Deutsche Bank's contract to manage SRO and SRQ, "in order to save what little is left", and intimated his willingness to engage in a proxy fight for control.

Since that time Horejsi has continued to accumulate shares (at latest count, 16.5% of SRQ), and he has come out in vehement opposition to the liquidation plan, calling it a dumb idea coming from "exceptionally bad management" and a complacent board of directors.

The DWS fund managers have responded vigorously. In order to keep Horejsi from blocking their plan to liquidate the funds, they have caused the funds to adopt "poison pill" defenses, to alter bylaws so as to allow arbitrary adjournments of shareholder meetings, and to rule out the selection of a new manager affiliated with a shareholder (i.e. Horejsi) unless 80% of the Directors approve.

These moves would be understandable if this were a typical proxy fight, with incumbents seeking to maintain the corporate entity and preserve control, but here the managers are fighting to keep control of the funds in order to obliterate them.

Why bother? Why is DWS so intent on having SRO and SRQ commit corporate hara-kiri that it feeds them poison pills to use in case anyone tries to rescue them? Why gin up the lawyers and the proxy tele-marketers, at shareholder expense, just to provide a quick death and burial for these unlamented funds.

It can't be due to concern for the reputation of the folks involved. Like superstar athletes on steroids, portfolio managers John Robertson, Jerry Ehlinger, John Vojticek and Asad Kazim have set records that will not be easily forgotten. Even in the market turmoil of late 2008, when the Vanguard REIT Index ETF (NYSEARCA:VNQ) was down by 39%, losing more than 90% of investors' capital in less than 3 months is the sort of accomplishment that should dissuade any conscientious investment firm from ever allowing any of them near any position of fiduciary responsibility, ever.

What other motive, then, can DWS have for insisting that SRO and SRQ put to sleep? Why should DWS care whether the books and records of the funds get recycled into pulp or fall into unfriendly hands? And just how does one manage to lose $200 million or $340 million in a few weeks, entirely through legitimate asset management, without any trace of Ponzi-like dishonesty and under a legal regime that demands full disclosure?

The proxy statement blames the losses on "unprecedented and intense volatility in the real estate market and increased investor fears and reactions related to the worldwide credit crunch". As SRQ's annual report put it:

The fund underperformed its benchmark and peer group by a wide margin mainly because of the fund's large leverage position, which badly hurt performance when investors began selling real estate assets indiscriminately during the early part of the fourth quarter as the worldwide credit crunch intensified.

Very true, as far as it goes. Both funds leveraged common stock with auction rate preferred. SRQ entered the 4th quarter with net assets of about $400 million: $160mm preferred (at face and redemption value) leaving $240mm net equity for the 15.7 million common shares, with a net asset value of $15.32 per share. SRO was a bit more leveraged: $730 million net assets backing $350mm in preferred shares, leaving $380mm for the 38 million common shares, or about $10.04 per share.

October was a devastating month, with VNQ (the REIT index ETF) falling over 31%, and November compounded the cruelty with an additional 23% drop. For geared funds like SRO and SRQ, such losses fall disproportionately on the common equity holders. The Investment Company Act of 1940 requires at least $1 of common stock net equity for every $1 of preferred. When net assets decline, this means that some preferred shares must be redeemed in order to restore the ratio, but this in turn requires funds to sell their holdings into a collapsing market in order to raise enough cash for redemption. SRO started the 4th quarter with roughly $668mm worth of investments and $53mm cash, but by the end of October it had only $231mm worth of investments left after raising an additional $201mm cash in preparation for a preferred stock buyback.

Going from $668mm to $231mm + $201mm implies a loss of $236mm, which fell entirely on the common shares, and is the main reason why net asset value crashed from $10.04/share to $3.61/share during the month. [Note: a loss of $236mm out of $668mm is a 35% decline overall, which is in line with VNQ's -31% for October. It meant a 64% decline for the common shares -- but that's leverage.]

However, a close look at the financial statement numbers suggests there's more to the story than just leveraged bad luck. Footnote A to each fund's Annual Report for 2008 says,

The Fund may enter into interest rate swap transactions to reduce the interest rate risk inherent in the Fund's underlying investments and issued preferred shares.

In fact, SRO entered the 4th quarter with five open interest rate swaps for a notional total of $437,500,000, under which it agreed to pay counterparty banks -- primarily the United Bank of Switzerland (NYSE:UBS) -- fixed rate interest at at 4% per annum, while receiving in exchange variable interest at the LIBOR rate. One swap had a 2008 end date, but the others extended as long as ten years. At the start of the 4th quarter SRO had an unrealized gain of about $2mm on these swaps, and by October 31st they were off by just $700K.

After October, though, the LIBOR interest rate fell dramatically, so the value of the cash flows SRO was to receive became much less than that of the fixed payments under the swaps. By year-end, four of the swap contracts had been wound up, at a realized loss of $20,604,055, and the remaining contract was under water by $12,929,062. SRO's swaps do not seem well matched to its portfolio or to its scheduled redemption of preferred shares. They look more like an interest rate speculation than the risk-reducing hedge described in the footnotes.

This speculation cost the common shareholders about $33 million during the 4th quarter -- almost as much as the $34 million of asset value that still remained for the common shares at year-end. The results at SRQ were not quite as bad: $11.3 million in swap losses were realized during the 4th quarter and the year-end position was only about $6 million under water, while SRQ still had $44 million in net equity remaining for the common shareholders.

The tax law says, in effect, that investment companies should invest in investments. Companies like SRO and SRQ are exempt from corporate income tax if at least 90% of their gross income consists of dividends, interest, securities lending, gains from selling stock, securities or currencies, and suchlike. (Tax Code Sec. 851(b)(2).) It seems, however, that DWS wanted to juice up the results for SRO and SRQ by having them buy into a privately held health resort, Canyon Ranch Holdings, LLC.

The problem is that part of the gross income from such a business shows up on the tax return of its owners, even if the net result is a loss, and this sort of income is potentially poisonous for the 90% test.

Indeed, SRO's liquidation proxy projects that the unlucky fund will fail to qualify as an investment company for year 2009, and Federal corporate taxes will have to be paid on this year's income before the balance can be distributed to the shareholders. [For SRQ it's still a close question.] Nor did Canyon Ranch fare well as an investment. SRO put $21,600,000 into it back in January, 2005. On September 30, 2008 it was on the books at $18,003,815. SRO's books show it was written down to $14 million in October, down to $4 million in November, and down to just $1,814,400 at year-end. It disappeared from the balance sheet entirely by the end of February 2009 -- whether by sale or write-off is not known. The 4th quarter loss on Canyon Ranch was $16.2 million, compared to ending equity of $34 million.

SRQ had a similar experience: Canyon Ranch went from a carrying value of $4.8 million at 9/30/08 to just $484,000 by year-end.

Canyon Ranch was not the only write-off. Both funds also owned preferred shares in hotel companies that had recently been taken private through leveraged buyouts. At the start of the 4th quarter, SRO valued its shares in Eagle Hospitality Properties at $3,586,700 and its shares in W2007 Grace Acquisitions I at $4,223,625. By year-end these interests had both been written down by more than 99%, leaving Eagle at $34,820 and W2007 Grace at $37,980.

One wonders if these illiquid shares still have any value at all, and how any rights that may exist can be enforced if SRO itself ceases to exist. For SRQ the story is the same: the value of its investment in these two companies fell from $7.1 million on September 30th to roughly $65,000 by year-end.

It's hard to know, though, whether to take these values seriously, because it's not clear whether DWS and SRO took them seriously. The "fair value" accounting rules say that funds like SRO should disclose how they value their holdings: are the numbers based on market prices (Level 1), other observable inputs (Level 2), or "significant unobservable inputs" (Level 3)? SRO's semiannual report showed Level 3 investments on June 30, 2008 were $39,041,135, which is the total of the values then assigned to Canyon Ranch, Eagle, W2007 Grace and a Rule144A investment in Hatteras Financial Corp. The year-end audited financials indicate that Hatteras had been transferred out of Level 3, leaving only $1,887,200 after write-downs as the remaining value of Canyon Ranch, Eagle and W2007 Grace.

However, the 3rd quarter schedule of holdings (issued between the other two reports, and certified, as usual, by the CEO and the CFO) shows Level 3 at $27,621,355, which equals the values then assigned to Canyon Ranch, Eagle and Hatteras. Something is missing. The investment schedule shows W2007 Grace being carried at $4,223,625, but it's not in the Level 3 total. SRO says that nothing was transferred into or out of Level 3 during the quarter, yet W2007 Grace was there on June 30, not there on September 30, and back there again on December 31. What kind of accounting magic is this? Maybe DWS employs double-entry gremlins or bookkeeping imps. And maybe those sprites also know whether the net asset values that SRO was reporting to the public in late November and early December did or did not include the $21 million loss that the fund had incurred on those interest rate swaps.

So where were the independent Directors of SRO and SRQ, the twelve "watchdogs" who are supposed to be looking out for shareholder interests? Four of them actually owned fund shares, though in nominal amounts. At DWS, however, a single consolidated board of directors is responsible for overseeing all of the closed-end and open-end funds in the DWS complex, so each Director of SRO and SRQ sits simultaneously on the Board of 131 or so other funds. A three hour board meeting, not counting coffee breaks, would allow about 81 seconds for discussing the affairs of any particular fund. And considering that each "independent" director collected upwards of $189,000 a year for their services to the DWS fund complex, it could be that the watchdogs didn't bark because they were too busy chewing.

As the Delaware Supreme Court said long ago about compliant Directors: "Whether they were supine merely, or for sufficient reasons entirely subservient it is not profitable to inquire. It is sufficient to say... that theirs is an unenviable position whether testifying for or against the appellants." Guth v. Loft, 5 A. 2d 503, 512 (Del. 1939.)

Coming from portfolio managers, excuses aren't much consolation for investors. Apologies would be nice, and compensation nicer still. But as the folks at DWS know: "dead funds tell no tales."

Disclosure: I still own the 500 shares of SRQ bought in happier times.