When looking back at the Great Financial Crisis of 2008, the primary catalyst the pushed the system over the edge and required central banks around the world to institute a global bailout of unprecedented scale was one simple thing: the layering upon layering upon layering of bets (using "other people's money" and courtesy of recently unleashed "financial innovation" in the form of virtually margin-free securities such as credit derivatives, demonstrated best by this chart) that interest rates would keep dropping, primarily in the form of exponentially tiered credit structures such as synthetic CDOs (all the way to the cubed degree) together with CDS sold on such layered synthetic derivatives. Of course, when the black swan event occurred and rates surged, this relentless leveraging of wrong-sided bets promptly resulted in the liquidation of any institution that was on the wrong side of such bets. Most notably AIG. In essence, AIG took the "logic" that since a rate blow up would likely result in the collapse of the US (and thus worldwide) funding structure, it would invoke the biggest central bank Put of all: either the Fed would rescue the world, or capitalism as we knew it would end.
As it turned out, AIG was right, and following the sacrifice of Lehman Brothers, every other institution on the wrong side of the levered "rate" trade was saved by the Fed. But at what price? Simply said, the Fed, in bailing out the world (a meme that has only now received popular acceptance following the release of formerly classified Fed documents, despite our claims precisely to that end from back in October 2009) has become the world's largest hedge fund and with a DV01 of over $1.5 billion by now, has taken on virtually unlimited interest rate risk (a topic discussed back in April 2010). As such controlling inflation expectations, or more specifically, Long-Term rates (the part on the curve that Quantitative Easing is powerless to control) is the most critical aspect of the viability of the monetary system. Stunningly, today we learn that to keep long rates low, the Fed may have resorted to nothing short of the same suicidal trade that destroyed AIG FP and brought the entire system to its knees. Namely, Ben Bernanke is now quite possibly the second coming of Joe Cassano, since in order to keep rates low, Bernanke is forced to a last resort action of selling billions upon billions of Treasury puts to "pin" rates low contrary to natural supply-demand mechanics. If so, the Fed is now basically AIG Financial Products, although instead of being synthetically long mortgages (and thus betting on a rate decline) and selling hundreds of billions in CDS to amplify its bet, Bernanke has done the same thing, only this time with Treasurys. Of course, Ben has the printing press on his side apologists will claim. Alas, that will have no impact whatsoever, if indeed the Fed has been reduced to finding ever fewer counterparties to a synthetic bet to keep long-term rates low, as very soon, with inflation ticking up, all hell may break loose in an identical replay of what happened to AIG once the Fed's put is called against it. Only this time there will be nobody to bail out the ultimate backstopper, resulting in the long overdue end of the current failed monetary system experiment.
Some may recall that over a year ago we made a curious discovery: by looking at the composition of securities held in the Fed's Maiden Lane I portfolio (than inherited from the collapse of Bear Stearns, which not even JP Morgan wanted) we uncovered that as part of the portfolio of toxic assets, which most recently was valued at $25.6 billion, the risk managed in charge of the book BlackRock had also put on a variety of synthetic hedges: "the FRBNY holds 5000 TYM0 puts, 3825 TYH0 puts, short 4000 FVH0, short 7828 TYH0, short 2240 USH0, and is short a bunch of eurodollar positions." The issue as we correctly specified, is that "while the Fed is pretending to care about interest rate concerns in an increasing rate environment and is hedging ML1, it has one billion DV01 risk for its house bailout package...This is a stunning number: the second rates commence creeping higher, you can kiss all that profit on TARP and what not not only goodbye, but the losses on the SOMA books will likely destroy America." We then concluded: "the Fed has decided to protect against a major hike in rates [in the Maiden Lane I portfolio]. Yet that which is truly relevant, the Fed's nearly $2.4 trillion in holdings of MBS, Agency and Treasuries is completely unhedged [the number is now $2.7 trillion and will be nearly $3 trillion by the time QE2 ends]. Good luck finding the counterparty that would be willing to put on a $200 trillion gross notional interest rate swap with the Fed." In other words, we were wondering why is the Fed not actively hedging its multi-trillion SOMA portfolio (including MBS, Agencies and Treasuries) if it was willing to do so with the far smaller Maiden Lane I subsegment of its holdings. Naturally, it may well have been doing so as there is no place in the Fed's weekly report (H.4.1) update that lists explicit derivative positions (more on this in a second). Ironically, it seems that we had the entire situtation backwards: it appears that far from being worried about hedging its SOMA book synthetically, the Fed may well have be constantly doubling down on its risk exposure in the form of off-book derivative contracts in order to "pin" Long-Term rates (read the 10 Year) by constantly selling Puts on Long Dated Treasurys at opportune times when there is no incremental buying of the underlying security, yet when, as the CDO and upcoming ETF debacles have so well demonstrated, the price of the derivative actually impacts the price of the underlying!
The missing sequential link in (lack of) logic comes from a report by Market Skeptics' Eric deCarbonnel who has combed through the June 24-25, 2003 FOMC minutes to find what could well explain the ongoing paradoxical flatlining in long-term rates even despite the threat of an end in QE2, which implies the removal of a buyer of some 83.4% of net Treasury securities, as well as the ongoing inflation threat so well described by James Grant earlier. What deCarbonnel has found is that as per then Fed secretary and economist, Vince Reinhart, and SOMA manager Dino Kos, the Fed has explicit authority and has in the past, sold puts on securities in order to bring various parts of the curve in line with "market expectations." The fragment from Dino Kos' transcript which implies that the Fed is likely actively pursuing a derivative feedback loop to keep long-term yields low (and thus prices high), is the following. Below, Kos discussed the "alternative approaches that would involve changes to how the Desk operates" in order to achieve the "conduct of monetary policy at very low short-term interest rates."
The alternative approaches that would involve changes to how the Desk operates are summarized in exhibit 4. The alternatives that could be adopted while changing only the composition of the balance sheet are listed in the top panel. These include (1) extending the average maturity of the outright holdings in the SOMA, (2) setting explicit ceilings on longer-term Treasury yields, and (3) using derivative instruments.
As deCarbonnel points out, 1 and 2 have already been either explicitly or implicitly utilized by the Fed in order to prevent the yield curve from exploding, due to the fundamental dichotomy of Fed operations: the Fed can keep short term rates at zero easily, it is the long-term ones that are a key threat to tipping the Fed's unhedged book over.
Which leaves only option 3: "using derivative instruments" to keep LT rates low.
And this is where it gets both interesting... and very disturbing.
Going back to Dino Kos' speech:
The Committee could sanction the use of various derivative instruments on conventional Desk operations as a way to influence longer-term yields, which is outlined in exhibit 8. Options of some form are a possibility, as are forward operations. For example, we could sell a sequence of options on term RPs, covering interlocking time segments that collectively extend as far into the future as desired. In this way, longer-term yields could be influenced and a visible signal of the Fed’s desired path of interest rates could be demonstrated. Forward operations in term RPs could be structured in a similar fashion.
And the stunner:
Alternatively, we could sell put options on longer-term Treasury securities at strike prices associated with desired longer-term yields. Of course, the operating objectives set for the sale of derivative instruments would determine their proper structure and should be carefully formulated first.
At this point the lightbulb should slowly be starting to glow:
The sale of any options, or forwards for that matter, would not affect the domestic portfolio immediately and, in the case of options, may never do so. Auctioning derivatives is something we already have experience doing. In the event that options were ever exercised, the impact on the portfolio would be profound, assuming that more than just a symbolic amount of contracts were sold. Simultaneously controlling the funds rate means that any reserve effect would need to be immediately sterilized. The volume of options sold might be limited because of this concern. Alternatively, options contracts might be configured to make a net cash payout if exercised, perhaps by structuring them as interest rate caplets or pairing them with offsetting trades with the Desk at then-current market prices. This would insulate the size and composition of the balance sheet, but the payouts would appear very visibly as losses on the income statement.
Summary: not only does the Fed admit that it has already sold off asset derivatives as a means of controlling short and/or long-term rates, but the Fed in essence is willing to do with rates derivatives what Warren Buffet did with equities in the form of his gargantuan index put sales, and Joe Cassano has done with CDS sales on his CDO holdings.
Here the Fed, as any rational investor seeking to manipulate the price of an underlying instrument, although with the benefit of having a printer, expresses the logical concern: what happens if options are excercised, or in other words, what might happen if the "pin" bogey on the underlying is crossed and the Fed suddenly finds itself in a losing "In The Money" position:
Of course, a successful program would be one in which any options sold would never be exercised. Achieving this result, just as with interest rate ceilings, would depend on how well the characteristics of the options—the strike price and the expiration dates—corresponded to market expectations for future rates.
Ironically this is precisely what Jos Cassano thought... Until of course Goldman changed the rules in the middle of the game, hiked collateral requirements and forced a toxic feedback loop whereby AIG had to undergo a liquidation waterfall putting it deeper and deeper underwater, until ultimately it was so far undercapitalized it had to be bailed out by taxpayers.
Kos logically realized that it is far easier to manipulate short-term rates than long-term and as such advocated initially merely dabbling in repurchase options, which only impact ultra-short term rates: i.e., those critical to bank functioning whereby banks can borrow cheap and lend rich.
In this regard, options on RPs with the Desk have a strong advantage over, say, options on Treasury yields because the policy rate over which the Committee has direct influence could be more directly linked to shorter term RPs than to longer-term Treasury yields. For these same reasons, options on Desk RPs could be structured to correspond directly with a policy commitment on the path of future short-term rates, and they could be effective through one of several channels.
We get even warmer:
First, even a relatively small program would undoubtedly add symbolic weight. Second, they would represent a monetary cost to the Federal Reserve of deviating from the implied path of future short-term rates, which might be seen as further binding the Committee to that path. For this effect, the more options sold the better. Third, a large volume of options sold could reduce risk premiums embedded in longer-term rates, independent of the level of credibility about any policy commitment. Here too, the more sold the more effective. As with interest rate ceilings, the question could be asked how effective the sale of options, either on Desk RPs or Treasury securities, would by itself be in reducing longer-term yields.
Kos' verdict: the Fed would need to sell a huge amount of Treasury puts to regain credibility that it would continue to sell even more puts should the situation require it: i.e., be the seller of only resort, and calm a Treasury liquidation wave by the market.
[The] ultimate success would hinge on the quantity of options sold—that is, how big a bet the Federal Reserve were willing to make. The more options sold, the greater the chance they would have the desired effect on longer-term rates even if not associated with any policy commitment, either by raising the costs to the Fed associated with options being exercised, or by lowering risk premiums on longer-term rates.
To be sure, Kos appreciated the downside risk associated with going all in on a losing bet, and then leveraging some more:
[O]f course the risks to the portfolio, to reserve levels, and of capital losses would rise in equal measure. And an exit strategy for options may not be as straightforward as it seems, even apart from the possibility of their being exercised. Of course, the Desk could stop auctioning new options at any time. But a decision to stop selling more options or not to issue new contracts with later expiration dates as time passes likely would be interpreted in the market as a statement about future policy intentions. The resulting rush to unwind market positions would likely be very disruptive and send yields sharply higher.
And that is the kicker. In essence the Fed may well be undergoing a program whereby via one of its Primary Dealers, most likely JP Morgan due to the banks key position as one of only two clearers of the repo system, it is selling Treasury puts, which would have an impact of pushing Treasury prices higher, and thus yields lower, contrary to all expectations in order to pin rates to specific levels. And as Kos admitted, the more long-term yields would run up, the more puts the Fed would be forced sell.
Since derivatives have little to no initial (or maintenance) margin requirements, especially not with a counterparty such as the Fed which can just print money any time there is a margin call, the Fed would be able to virtually print an endless amount of Treasury puts to keep underlying, and very much delta hedged, position, read THE YIELD ON THE 10 YEAR precisely where it wants it! The Fed says as much in Exhibit 8 to the June 24-25, 2003 minutes:
And before skeptics say the Fed would never do this in reality, Vince Reinhart admits that the Fed did very much that just over a decade earlier:
The System has also been willing to put its balance sheet at risk to encourage appropriate expectations about interest rates or to calm fears about funds availability. As plotted at the top right, the Desk sold options on RPs for the weeks around the century date change that totaled nearly $0.5 trillion of notional value. Given that the Desk already operates in all segments of the Treasury market, we wouldn’t have to move up a learning curve if instructed to increase purchases of longer-dated issues.
With lack of data availability one can only speculate how much options, most likely in the form of swaptions, the Fed would need to sell currently to keep 10 Year yields low, although if past is any indication, if the SOMA desk sold nearly $250 billion in repo puts back in 2000 when the Fed's balance sheet was a fraction of what it is now, it is safe to assume that a comparable amount currently would have to be in the trillions, if not tens or hundreds, considering the far lower notional impact on a security with material duration compared to one to impacted (and impacting) by merely ultra-short term rates.
As for the Fed's justification to impact the market in such a covert and off-balance sheet manner some might say, it provides the following justification:
The Federal Reserve has always appreciated the importance of correctly aligning market expectations about the economy.
Of course when the Fed sees the economy as the market, such as now, it is critical that the Fed do all in its power to prevent an efficient price discovery process from occurring.
For those for whom this is still unclear, basically what the Fed may well have done (and has admitted to doing in the past) is what the market does each and every day, when ETF buying results in a long/short gamma trade that pulls or pushes component securities higher or lower. There is a reason why the SPY and the ES are the two most liquid securities in the market: control these, and what happens to the underlying stocks is irrelevant. In very much the same way, the Fed which still continues to load up on Treasury securities (albeit far less so as the longer-end of the curve), is now most likely pursuing a goal of keep the curve as flat as possible and not losing the long end.
As a reminder, during QE Lite/QE 2, the Fed has practically forsaken the 30 Y sector of the curve, or the part the most reliant on long-term inflation expectations. Why would the Fed do this in the cash market unless it had some other way to definitively impact demand via synthetic instruments?
A synthetic off-book short put position also explains the confusion of those such as Bill Gross: when the Fed supposedly exits the monetization game, regardless for how short, rates would traditionally be expected to rise. Yes... but only if the Fed was not concurrently selling massive amounts of volatility. And while the actual buying and selling would remain hidden from public view, the aftermath would be visible in downstream market effects. Indeed, this is precisely the case. As can be seen on the below chart which maps the yield on the 30 Year during various QE regimes, and the level of the MOVE Treasury volatility index, during times when the 30 Year appears poised to break out higher in anticipation of a QE end, yet merely trades rangebound, the level of MOVE plummets.
Obviously, plunging vol ends up having an offsetting impect on prices. Were the vol not to drop so materially, the upward drift in Long-Term yields would be very pronounced, and result in the 30 Year breaking out north of 5%, after which the Fed would likely lose all control of the curve. The question then becomes: who is selling all this vol ahead of such a risky event as the end of a Quantitative Easing episode. Our bet: none other than the current Manager of the System Open Market Account: Brian Sack.
This however leaves open the question of just where on the Fed's books would Sack et al keep a record of how much the Fed has in Treasury yield exposure?
Enter Exhibit A: Federa Resere "Other Assets", which as we have been disclosing for quite some time is now a not too negligible, and very much record, $125 billion! That's right, the Fed has $125 billion in other assets, whose definition is so broad they could well be anything and everything, and which we have a sinking suspicion could well include among them the "net" exposure the Fed currently is booking on its swaption book.
Naturally, the specific booking of these "assets" would depend on the format under which the Fed has contrived to make its synthetic but on higher Long-Term prices: whether these are outright puts, or, as was disclosed previouly when looking at Maiden Lane I, in the form of curve swaptions (and thus payers or receivers).
Recall that as per the latest breakdown of Maiden Lane I assets, the actively managed book includes, among other things, $675 million in net notional swaption exposure (gross could be anything), as well as ($3.3) billion in Interest Rate Swaps. Is this the same open market tactic that the Fed is applying to the broader Treasury curve? If so, keep a very close eye on abnormal activity in the swaption market, especially that originating from JP Morgan.
Which leaves open one question: with the Fed selling Treasury Puts, who is buying them? Well, the answer could be anyone. While nobody in their right mind would every transact directly with a Fed selling insurance on its own books, many Primary Dealers, and foreign institutions are certainly eager to hedge their surging Treasury exposure, courtesy of the trillions in new issuance each and every year. In this scenario, JPMorgan, the Fed's proxy, would all day, every day, especially on key inflection date such as ahead of the termination of a quantitative easing period, flood the market with an overabundance of Long-Term volatility (i.e. selling puts), which as shown above, would push the MOVE index lower, and result in a strengthening of 10 Year rates.
Yet what observent readers will realize is that those buying protection from the Fed would be analogous to those who would have been buying protection from AIG on its CDO books. However, unlike back then, when entities like Goldman were sure to be bailed out on their collateral exposure to Joe Cassano, this time around if the Fed does lose control of the long end, and the Fed ends up experiencing first tens, then hundreds of billions of P&L losses should rates jump by 1%, 2%, 3% or more, then all bets would be off, and everyone left holding Treasury protection, i.e., the bet opposite to that of the Fed, would end up with with a big, fat nothing as the monetary regime finally fails, fiat is dethroned, and alternative monetary systems are implemented.
If indeed the Fed is the primary driving force behind the long-end's continued resilience, one should first inquire whether or not this is an action that the Fed is misrepresenting as performing - after all in its description of activities performed under US Foreign Exchange intervention, the Fed clearly states "The Fed historically has not engaged in forward or other derivative transactions" - is this blanket statement true only for Fed's FX intervention regime or for everything, because as Reinhart has confirmed, if so, the Fed is engaging in activity that is not previously disclosed as performing. Second, if the bulk of Treasury put buyers were to realize that their ultimate counterparty, under the guise of various Primary Dealers, and especially JP Morgan, is indeed the Federal Reserve, they will promptly abandon the Treasury Derivative market, forcing the Fed to lose this key lever of reverse market influence, resulting in chaos when it comes to controlling the long-end, and leaving the Fed with a curve that is at or near zero on the short end and surging in the 10 Year and over spot, which would result in complete loss of control by the Fed regarding inflationary expectations, the collapse of the dollar (yes, even more than to date), and an explosion in commodity prices as the nation finally careens over to its Weimarian endspiel. Lastly, a key question to demand of Bernanke, if it is confirmed that the Fed is shaping the yield curve using derivatives, would be just how great the Fed's blended risk is currently over and above the DV01 on merely its underlying physical instruments, and just how the Fed will hedge not only its massive ~$3 trillion paper exposure, but possibly its multi-trillion synthetic exposure should inflation surge, and the Fed's SOMA desk finally lose control of the long end.
While there are many questions embedded in the assumptions presented above, we are confident at least some unconflicted reporter will inquire Ben Bernanke on April 27 whether any of the above is in fact true.Which if confirmed, will be the biggest admission of market manipulation by the Federal Reserve in history.
Those who wish to follow deCarbnnel's thoughts on the matter can do so below...
And for those curious to see just how the Fed predicted this very course of action as long as 8 year ago, we urge you to read the following exhibit from the Fed June 2003 minutes in its entirety - link
What is most ironic, is that Mr Reinhart was kind enough to leave us with the Fed's next steps for when the SOMA manager finally loses control of the curve. From Exhibit 8 to the June 2003 minutes (presented below in their entirety):