Magnetar Capital LLC 847.905.4400
1603 Orrington Avenue 847.905.5868 fax
Evanston, Illinois 60201
April 19, 2010
Dear Magnetar Capital Investor:
We consistently strive to maintain clear and open channels of communication with our
investors. In pursuit of this goal, we wanted to address the recent publicity regarding our
Mortgage CDO investment strategy. At the center of these stories is a blatantly false and
misleading story written by ProPublica, an online news outlet, regarding our Mortgage CDO
investment strategy that was active from 2006 through 2007. Despite our best efforts to
educate ProPublica’s reporters about the specifics of Magnetar’s Mortgage CDO investment
strategy as well as the general process and market circumstances regarding the structuring
and issuance of CDOs, ProPublica simply got the story wrong.
This letter is intended to address several of their inaccuracies, but more importantly to
provide you with a recap of the pertinent facts regarding our Mortgage CDO investment
strategy. To that end, we have detailed herein much of the same information that we had
discussed with these reporters prior to publication of their story.
Assuming that the reporters understood the information that we relayed, they chose to
disregard the key fundamental facts that we outlined in our discussions and supporting
documentation. ProPublica claims that Magnetar invested in CDOs that were built to fail
(and preferably to fail as quickly as possible so as to maximize returns) in order to
capitalize on a fundamental view that the US subprime mortgage market was going to suffer
significant defaults. Since it first published its article, ProPublica has been continuously
retelling its story regarding Magnetar as if it were fact. Ironically, there may very well be
other market participants whose strategies are accurately described by the ProPublica
thesis. However, that does not make it true in our case.
As we explained to the reporters, and have explained to you, our investors, consistently
since the early part of 2006, our Mortgage CDO investment strategy was based on a
mathematical statistical model. There was no embedded view regarding the direction of
housing prices, the rate of mortgage defaults, or the subprime mortgage market generally,
and, therefore, the strategy was designed to generate and maintain a market neutral
portfolio over time. In its simplest form this strategy was a capital structure arbitrage. Our
models simultaneously took into account all CDO equity tranches in which Magnetar
invested, together with appropriate levels of portfolio hedge positions, in a global
framework and not on a deal‐by‐deal basis. Our portfolio construction was designed to
make attractive returns over a long duration (an estimated average portfolio life of 6 to 8
years) if the housing and mortgage markets were healthy, but contained a prudent level of
hedges to provide for the performance of the strategy in a continuum of scenarios in which
the overall housing and mortgage markets might lose value. It was not a “bet” that any CDO,
any group of CDOs, or the housing or mortgage markets as a whole would fail either in the
short term or the long term.
We do not feel it is constructive, or in some cases even possible, to address all of the article’s
assertions and speculations, or the variations on ProPublica’s theme that have been
expressed by other media outlets. We realize that many of you are likely to get inquiries
from your own investors and colleagues on this article and we wanted to ensure that you
had the detailed information needed to answer these questions accurately. For many of you
these points, and the graphs illustrating them, should be very familiar, as we have not
varied in the description of our Mortgage CDO investment strategy from its inception in
Our Mortgage CDO investment strategy was based on a market neutral statistical
framework which specifically did not incorporate a fundamental view regarding the
direction of the housing or subprime mortgage markets.
Magnetar’s portfolio varied over time, but both the initial construction and subsequent
modifications of the portfolio were at all times managed to be market neutral. The
assertion that Magnetar built its investment portfolio with the hope and/or expectation that
CDOs would fail either in the short or long term, is simply inconsistent with the
construction of our portfolio.
As we stated earlier, Magnetar’s strategy was in essence a capital structure arbitrage. This
type of strategy is broadly employed in corporate credit markets, and is based on the
relative value between differing components of a company’s capital structure (in our case
the different tranches or classes of a CDO), and on the supply‐demand imbalances which can
be exhibited in the pricing of rated and non‐rated tranches. From early 2006 to late 2007,
there was a systematic relative value mispricing between the equity tranches of Mortgage
CDO structures, which offered approximately 20% target yields, and mezzanine debt
tranches of Mortgage CDO structures, on which credit protection could be bought for
between 1% and 4% (depending upon which tranche and CDO). The facts detailed below
regarding Magnetar’s execution of this strategy contradict ProPublica’s speculations.
Magnetar’s investment strategy was not based on a desire to construct deals with “riskier
assets” so that we could place “bets that portions of [our] own deals would fail”.
• From the inception of the strategy through to its conclusion, Magnetar believed that
its overall portfolio, including macro hedges, would be profitable independent of the
direction of the housing and subprime mortgage markets. We clearly explained this
fact, and the reasoning behind it, to ProPublica. The graph below (Graph1)
represents the average portfolio payoff profile over time, and clearly demonstrates
that the payoff profile in respect of CDOs in which Magnetar invested was long‐
biased. Of particular note is the lower, darker line which represents the payoff
profile of our long investments in these CDOs plus only those hedges referencing
tranches in these same CDOs (that is, excluding the performance of hedges we
purchased that related to transactions in which we had no long interest).
Note: Due to considerable trading activity, the graph (Graph 1) above is an approximation but certainly
is representative of the portfolio.
• The expected return (unhedged) on the equity tranches of the CDOs in which
Magnetar invested was estimated to be approximately 20% per year if the housing
and mortgage markets remained robust (and, accordingly, the underlying assets
performed well). Magnetar purchased hedges (out‐of‐the‐money credit protection)
on these same CDOs with an estimated cost of less than 3% per year. This is shown
as “CDO Hedge Cost” in the graph (Graph 2) below. Magnetar purchased additional
hedges (credit protection) referencing CDOs in which Magnetar had no equity
ownership interest, as a macro hedge to insure its overall portfolio was market
neutral and did not express a view on the direction of the market. The expense of
this additional hedge was again less than 3% per year. This amount is represented
as “Macro Hedge Cost” in Graph 2 below.
• As you can see in the CDO Portfolio Analysis graph below (Graph 2), only a portion
of the income created by our long CDO investment strategy was utilized to finance
the cost of the hedges on the portfolio. Had our goal been to create a short bias in
our portfolio, the hedge ratio should have been approximately 5 times greater just
to arrive at a net carry of zero.
All of this stands in stark contrast to assertions made by ProPublica and others, that the
strategy would only be profitable if the CDOs in which Magnetar invested failed on a grand
scale, or that we had constructed a portfolio in which the yield on the long positions would
be barely sufficient to cover the cost of carrying the hedges and was nothing more than a
clever self‐funding mechanism.
Magnetar did not control asset selection in CDOs in which it participated.
The authors of the ProPublica article have also asserted that Magnetar was in control of the
asset selection of the CDOs, or at the very least pushed for the CDO managers to include the
riskiest assets in those deals. Magnetar at all times only acted as the equity investor, not as
the manager of a transaction (the “Collateral Manager”) or investment bank arranging the
transaction (the “Dealer”). As the equity investor, Magnetar often communicated with the
Dealer and Collateral Manager during the transaction process. Magnetar, in its role as
equity investor, did not select or have control over the assets that went into a CDO. A closer
look at the asset selection and marketing processes of a transaction highlights both the
limitations of Magnetar’s rights as an equity investor and the ultimate checks and balances
the Collateral Manager and the Dealer had on the asset selection process.
The Dealer and a Collateral Manager would agree on the general terms of a proposed CDO
transaction. The Dealer and Manager would generally receive market feedback from
potential investors including, most commonly, Super Senior buyers and equity investors
(such as Magnetar) to assure that the general terms of the proposed deal would be
marketable and, if possible, to secure “lead orders” for the transaction. Magnetar typically
provided an equity purchase letter that unambiguously described the general terms upon
which Magnetar would be willing to invest, including the requirement that assets would be
selected and managed by the Collateral Manager. In the typical CDO, over the 3‐5 months
following the initial agreement between the Dealer and the Collateral Manager, the
Collateral Manager would select assets and direct the Dealer to purchase and inventory the
assets prior to the closing of the CDO (the “warehouse” period).
The Collateral Managers relied upon their investment processes to choose the assets for the
CDO. The Dealer, in each case, had to affirmatively agree to each and every asset purchase
during the period that the assets were warehoused prior to the closing of the CDO and the
issuance of securities. When the Dealer purchased an asset and put it in the warehouse, it
took risk on its balance sheet. The Dealers’ requirement to approve the inclusion of every
asset in the portfolio prior to the closing of the CDO was a protection established for the
Dealer. It allowed the Dealer to review the quality of the assets to reduce the risk of
including an asset that would cause a loss while it was on the Dealer's balance sheet. The
process also allowed the Dealer to monitor the creation of a portfolio so that it would be
attractive and acceptable to investors once the marketing phase commenced.
Once a substantial amount of assets had been purchased into the warehouse, the Dealer and
the Collateral Manager started their marketing effort and presented the proposed
transaction to potential qualified investors. Each potential investor (whether a potential
investor in the equity or the rated debt securities) had the opportunity to fully analyze the
portfolio, discuss particular assets with the Collateral Manager and challenge any of the
manager’s selections before making an investment decision. If there were an insufficient
number of qualified investors that chose to invest or subscribe for debt or equity tranches
of the CDO, the Dealer would either be left with the unsold interests in the CDO or all of the
The process described above was followed consistently for each of the mortgage CDOs in
which Magnetar was an equity investor, although the process for other CDOs in the market
might have been different. In particular, it should be noted:
• Magnetar did not participate in the marketing or distribution of CDO transactions.
• Magnetar did not make any representations to qualified investors who purchased
• On average across Magnetar’s entire portfolio, less than 7% of the aggregate assets
of these CDOs consisted of CDS where Magnetar held a hedge position in the same
instrument. While there was significant variation among the CDOs in which
Magnetar held equity, this number was derived from the following conservative
o While Magnetar was an active purchaser of CDO protection (hedge
instruments), in all cases Magnetar’s hedges and its long investments were
transacted with a Dealer as counterparty.
o Because Magnetar’s transactions were always with Dealers and never with
CDOs, we are unable to know with certainty how many of our protection
instruments ultimately became assets of the various CDOs in which
Magnetar purchased equity.
o However, for purposes of the statement in the principal bullet above, we
have assumed the most conservative possible outcome (that is, we assumed
that all of Magnetar’s protection instruments became assets of the various
CDO’s in which we held equity, to the extent that the assets in the CDOs’
warehouses corresponded to the protection instruments held by Magnetar).
Magnetar’s portfolio and structural preferences were expressed in a manner
consistent with general market practices.
ProPublica asserted that Magnetar requested higher spread collateral in the CDOs in which
it purchased equity, for the purpose of increasing the likelihood that a “bet against” the
deals in which Magnetar was an equity investor would become profitable due to the failure
of the deal. They posted the email discussed below as support for their assertion.
However, the context of that email actually shows that Magnetar’s intentions were to
increase the attractiveness of a potential “long” investment in that deal, by bringing the deal
in line with then‐current market norms. The CDOs of the type in which Magnetar purchased
equity interests were specifically synthetic Subprime Mezzanine CDOs. The market rate of
return that was typically expected by equity purchasers of these CDOs at that time was in
the range of “mid teens” to “low twenties” depending on the deal. Magnetar’s CDO
transactions were conducted within this range of expected returns – they were market
In order to achieve these market rates of return and actually close a deal that was attractive
to potential purchasers of CDO equity, the portfolio of assets underlying the CDO required a
certain level of yield. As we noted above, there would typically be ongoing interactions
between potential investors (including Magnetar), the Dealer and the CDO Manager during
the warehouse period. This was part of the process by which the investors would satisfy
their diligence requirements and assure themselves that the CDOs assets would have the
appropriate characteristics for a transaction of this type.
ProPublica has asserted that Magnetar pushed for specific assets to go into specific CDOs. It
published the email referenced above and claimed that the email supported this assertion.
However, ProPublica intentionally failed to publish the accompanying attachment to that
email, which we provided to them, that made it clear that Magnetar’s sole interest was in
achieving a portfolio with certain broad characteristics. The email did not mention or
suggest individual assets. The full email and its attachment are included as Appendix A.
The following facts are particularly noteworthy:
• The email relates to a transaction that was proposed but never completed.
• Originally, the Collateral Manager proposed an overall portfolio weighted average
spread (WAS) of between 101 and 136 basis points which was well below the
typical market levels for subprime mezzanine CDOs. Magnetar estimated the
proposed transaction would produce returns on its equity investment of between
1% and 7% if the portfolio suffered no losses. This transaction was, therefore, not
attractive either on an absolute basis or in comparison to similar transactions in the
marketplace at the time. Accordingly, Magnetar responded and requested a “higher
spread” portfolio that could earn the equity holder market level returns,
commensurate with the perceived risk and more comparable to other similar
transactions in the marketplace.
• As you can see from the attachment at Appendix A, Magnetar’s response proposed a
different sector composition with wider asset spreads. Magnetar’s behavior was
consistent with a rational equity investor’s interests and with a process which was
standard not just in the CDO market, but in investing generally.
• The same Collateral Manager subsequently executed a CDO with Bear Stearns in
December 2006 which had a minimum spread requirement of 188bps. This
transaction was initiated shortly after the email discussed above was written, and
was completed at spreads approximately 10% wider and with a weighted average
rating factor (or WARF) 7% higher (and implying a need for commensurately
“riskier” assets) than what Magnetar had proposed in its email. Magnetar was not
an equity investor in that subsequent transaction.
• As has been discussed with many of you in the past, the “triggerless” equity
structure that Magnetar preferred for its long investments required a significantly
larger initial equity tranche investment, increasing the required long investment by
at least 50 percent for a given CDO portfolio. If Magnetar had a fundamental short
view, with an expectation that the equity would be written off, it would have been
irrational to purchase equity in a triggerless CDO instead of taking that extra 50
percent investment and using it to buy additional credit protection. Even the
argument suggested by ProPublica and others, that our longs were funding our
shorts, would not require us to acquire (and risk) anywhere near the amount of CDO
equity that our portfolio held relative to its hedges.
We hope that the above provides you with more detail to clearly understand our role and
actions in the marketplace. And we further hope that having reviewed all of the facts, you
will be able to state with confidence to anyone who asks, that Magnetar’s strategy did not
engage in the “Magnetar Trade” as coined by the media. We do not anticipate commenting
on additional media stories, as our priorities remain focused on managing our investor
Magnetar has been built on “best in class” investment processes and the business,
operational and legal infrastructure required to support them. We remain committed to
serving you in our capacity as stewards of your invested capital. We truly appreciate your
continued trust in and support of Magnetar.
Magnetar Interaction with Manager
Dealer Interaction with Magnetar and Manager
Magnetar Interaction with Manager and Dealer
Portfolio Composition as Proposed by Dealer
Portfolio Composition as Proposed by Magnetar
Magnetar Interaction with Manager