Treasury "Fails" Revisited......

Today, I’d like to do something a little different than usual.  In the context of my recent “$128.9 Billion of Failed Treasury Transactions last week......and My friend Harvey"  post, I wanted to take some time drilling down on this phenomenon and get an expert opinion on what might really be happening here.  I wanted to find someone who eats, sleeps and breathes this stuff.

To that end, through a mutual friend, I got in touch with someone, who I consider to be an expert on the bond markets.  I asked that he take some time to guide me though what seems to be happening with these “fails”.  He agreed to be interviewed and I’d like to share his insights with you in this post.  I’m referring to Tim Bergin.

Tim’s background is that of a proprietary credit trader.  He worked at banks and hedge funds trading and managing portfolio’s that consisted of corporate bonds, convertible bonds, credit defaults swaps, equities, and other derivatives.  He did that for 12 years in New York and Toronto before leaving to manage his own money and provide ideas for clients.


Tim Bergin Interview:

Deep-Throat-IPO (Me): 

Tim, we’ve had some time to e-chat a bit on what might be causing the “fails” or disruptions in closing these transactions.  The numbers have been growing quickly, as you know.  In October, roughly 30% of the transactions “failed”.  Now, over the last few weeks, more than half of the transactions are failing.  (weeks 11/22 through 12/20 - $1.366T/$2.443T = 56%).  Keep in mind that transaction failures ran at only 5.1% in 2012 and 9.2% in 2013.  As I mentioned, I’m concerned that we might be heading for a Lehman moment.  Can you give me your thoughts on this acceleration?

What’s up?

Tim Bergin:


Hi, thanks for the kind words.  While I traded many (billions) of US treasuries over the years I was more focused on corp bonds, convertible bonds, and credit default swaps.  So I could be wrong on some of the treasury market specifics, but I think I can shed some light on the bigger picture issues.

Quick answer, it is not as bad as it seems.  Due to regulatory changes the amount of bonds that can be posted for collateral has shrunk.  The regulators really want high quality collateral such as US treasuries to be used in repo financing.  Mandates have changed requiring banks, insurance companies, etc. to hold more US treasuries.  They hold more treasuries to satisfy  liquidity ratios and capital requirement changes favour such assets.  There has also been more assets allocated to treasuries through etfs, pension funds, sovereign wealth etc.  So what you can see is that demand for treasuries has increased and the demand is from entities that don't lend out their portfolio.  Add to that QE, central banks from around the world buy US treasuries, and again don't lend.  You can see there is a huge imbalance between demand for US treasuries to use in repo and the supply.

These effects have meant that most of the liquid US treasuries, you would want to be used for repo, trade 'special' meaning they are very expensive to borrow.

An example: a typical credit trade would be the purchase of a corporate bond versus a short position in a similar maturity US treasury.  We would often find that it was getting more and more expensive to be short the liquid ‘on-the-run’ US treasuries, as the borrowing charges were increasing because of the effects described above.

So my guess with treasury fails is that dealers are left net short US treasuries and can't find them to buy to settle the repo trades.  It is not an economic risk as they can offset this short  10yr on-the-run bond with a long in a 10yr off-the-run bond that doesn't have the same excess demand.

This is another example of the distortions central banks have created, but not a sign, at least I don't think, of funding pressures.

One other part to add is that the person that is holding the collateral in the repo is fine with the transaction failing.  While they are in possession of the 10yr bond they can earn extra money lending it out (taking advantage of the 'special' rate).  So they aren't really fussed about the repo failing.  This issue is well known.  Central banks are aware, and aware that they are largely to blame.  If I am right this issue will go away as more treasuries are supplied (especially when new tax laws go into effect) and with QE ending.

I really don't think that funding is a concern for US banks and brokers.  Take Morgan Stanley, their repo is only $54bil, which is small relative to total liabilities of $773bil.  The issue with Lehman was that they were pledging CDO's and garbage like that as funding instruments.  When the collateral would no longer be accepted by other dealers then the funding issue started.  Repo issues are more a problem when the collateral itself is the issue.  With rules that stipulate that mostly treasuries need to be used, I really don't see this being an issue.


Deep-Throat-IPO (Me):  

So, to paraphrase, you think the failure to deliver or receive the securities is more of a “matching” problem.  The parties intend to complete the transaction, but for various reasons, the specified security isn’t available and the transaction fails.  The counterparty is fine with it failing because they have little or no economic cost.  They either have “the money” if they fail to receive the collateral, or other collateral if they fail to deliver, and can put either to good use until the “matching” problem is corrected.  If I understand you correctly that makes some sense.

You also mentioned that you don’t think that there’s currently a funding problem, but I’d like to dig into that a little more.  I read the occasional article or two about hedge fund & trading risk built into the system.  I.e.) A money manager applies leverage to goose returns on “safe” securities, since his investors are looking for “safe” returns of say 6%, unachievable in today’s environment.  But if they apply a little cheap-money leverage, say 3:1 or 4:1 on a treasury transaction, the returns are suddenly within reach and things look pretty rosy.  The way I see it, there’s some risk if there’s ever, as you folks say, a “gap up” (yields rise and bond prices fall) because of some forced selling.  Quoted prices are always accurate until someone has to raise cash.   Can you talk a little bit about liquidity?  Can you sell what you want to sell at a fair price right now?  …..or do the Bloomberg screens suddenly “go dark” when big sell orders start showing up?….do you see any other Lehman-esque “garbage” out there right now?  Are there enough buyers out there?

 Tim Bergin:


A gap up in rates is my biggest concern for markets.  While everyone thinks equities are expensive, and they are, I really think the mispricings in bond markets (particularly in Europe) are much, much worse.  The amount of leverage in the system exacerbates this issue much more than people realize.

You are right about leverage, before the leverage in the system was at the banks and brokers, now the leverage is with non-bank market participants.  As yields have dropped, client return expectations haven't.  So if you are a bond manager at a hedge fund and rates are at 2%, your clients still expect 8 to 10% returns, what do you do?  As you note, the answer is leverage.  Plenty of these guys have been buying up front end short maturity debt and levering 5x or more. 

Another group using large amounts of leverage in bonds markets is ‘risk parity’ type strategies.  Their models, I believe, encourage more leverage as yields drop. 

This leverage means that a large move higher in treasury yields could create a lot of selling pressure from these highly levered entities, creating a feedback loop of ever higher yields.  It would be a similar to what some people forsee in equity markets with the crowded long vix trade.  It can create a selling feedback loop if equity markets dip past a certain point.

In the scenario I am describing, it would be a hedge fund with $1bil in funds that borrows money to buy $5 to $8bil of fixed income assets.  Now to take a step back for a second, one thing people don’t realize is that as interest rates have dropped the interest rate sensitivity of bonds has increased.  The previous rule of thumb was 8% losses per 100bp more in yields for a 10 year bond.  However, things are different now.  The math has changed.  To explain, if you have a 0% coupon bond the sensitivity to rates is the same as the maturity of the bond.  As yields have dropped close to 0% the sensitivity has then increased.  So now the 8% is more like 9%.  Doesn't sound huge, but people are now more than 12% exposed to a move in yields than they used to be.  This is much more pronounced in 30yr bonds (like your favour Baba bonds).  The sensitivity has gone from 18% per 100bps to something like 25%.

Now most hedge funds employing such a strategy do prefer shorter dated bonds, which is safer to apply leverage too, but certainly not all are being prudent.

So to summarize to this point.  More people are using higher amounts of leverage in fixed income markets to achieve returns.  The interest rate risk, per unit of bond, is the highest it has been in history.  Sounds like a good mix for a blow-up, right?  It gets worse.  Bond market liquidity is awful.

When I first started in the business back in 2005, the standard trade size for a corporate bond could easily be $25 to $30mil.  Call Goldman Sachs, JP Morgan, or UBS etc, ask for a market and sell to them.  A clip of $100mil in corps could easily be done.  With Volcker rule that has stopped.  Perhaps the level of liquidity provided in 2005 was too generous, but the pendulum has swung too far and conditions now are really bad.  I would guess that most corp trades are $2 to 5mil.  Before you could just see a price on screens and transact, now dealers want to work orders.  Meaning they will indicate a market where they think the bond should trade, but won't buy for their own balance sheet at those prices, instead they try to find a willing buyer at the price indicated.  If you are Pimco or Blackrock the experience would be much better, but for most market participant its tough to trade in meaningful sizes.  Optically, to outsiders the bond trading world doesn't look so bad as bid-ask spreads on Bloomberg screens don't look that wide relative to history.  However, those Bloomberg screen markets are now mostly just indications.  You can't just transact at those prices like you could in the past.

The huge question is how does that market function in a period of forced selling?  There has also been a rise in corp bond etf's that would also become a seller in this environment.  Adding up all these risks and you can paint a scenario where bond markets get really really ugly very quickly.

Will rates gap higher?  No idea.  But the new tax changes, recent signs of inflation, less QE, and possibly strong economic data forcing central banks to raise rates means that the probability of such an event have increased.  The chances of ‘something’ happening are definitely higher.

I don’t see too much ‘Lehman garbage’ but I think there is a huge amount of market complacency to the amount of underlying risk there is in the bond markets.  To give some credit to the market, being complacent, riding the momentum, and assuming you do have a central bank ‘put’ (if markets get ugly) has been a phenomenally profitable trade the past 8 years.

However, the real systemic risk is in the knock-on effect of such a disorderly bond market.  This is something I think a lot about.  Higher yields could create solvency issues.  Banks and insurance companies are all very long bonds (often due to regulatory mandates) and all are levered entities.  A gap higher in yields as described will create a big loss/equity impairment at all these institutions.  But it’s not just bonds.  Real estate assets are very rate sensitive and the primary collateral for lending.  So you could see a scenario where there a large losses on investment portfolios at banks, but then also the increase in credit risk as the collateral against loans drops precipitously at the same time.

I know of a few REITs that have their property portfolio marked at 4.1% cap rates.  At 7% cap rates (cap rates should jump in concert with higher yields) I believe the assets would be worth less than the bonds outstanding.  Meaning equity is completely wiped out.  I could imagine there will be plenty of other situations like that as well. 

Obviously you can argue that the Fed steps in if things get this bad.  That very well may happen, but then you have to ask yourself the question, are we now forever confined to lower rates?  Some people think so, but I don't.

To summarize my concerns, so much asset value is tied up in low rates, and there is a chance that rates make a violent move higher triggering huge losses.  Both due to leverage and due to asset price reductions.

This scenario may never come to pass but it is the one that I think and worry about the most.


Deep-Throat-IPO (Me):

Well!!….Happy New Year Tim!......after thinking your commentary through I have pretty much decided to trade my silly, sequined 2018 Party Hat in for a crash helmet……LOL

But before I do, let’s dig a little deeper into the liquidity thing.  Here’s your comment from above:

"The interest rate risk, per unit of bond, is the highest it has been in history.  Sounds like a good mix for a blow-up, right?  It gets worse.  Bond market liquidity is awful.”

I remember reading sections of Michael Lewis’s book, “Flash Boys” where Brad Katsuyama first discovered the HFT effect.  He would put larger (albeit equity) trades through and they would only partially execute at his price, and as I recall, the market would literally “disappear from the screen”.  It sounds like you’re saying that the “fake” liquidity in the bond market is a byproduct of a lack of security availability (not enough participants at the quoted prices)  i.e.) there’s probably enough liquidity in the system, but the bids/asks are fake/illusory and not applicable to larger orders….. giving us little guys at least some level of confidence that all is well, while simultaneously providing an opportunity for some enterprising folks to skim a few crumbs off the cake?

I think about my on-line brokerage accounts and I assume that on any given day I can sell whatever I want to sell at something close to the price I see on the screen.  With the “feedback loop” and possibility of the “Gap-Up” you’re describing, it sounds like, specifically in the bond markets, there may very well come a day where that’s not likely.

That said, I also want to make sure I (and our readers) understand the math.  As interest rates declined after the crisis, the higher coupon debt gained a ton of value (the 9% rule of thumb).  People who knew what they were doing got rich.  Wouldn’t it have been great to have loaded up on those 6% Aaa Corporates in 2006/2007?  Perhaps borrowing a few bucks to do it?

And now, just to throw a few gigantic numbers around (I really enjoy Gigantic numbers).  Per the BIS, latest figures (Q2 2017), there’s roughly $91 Trillion of rate sensitive securities out there …..$39T are US issues, $12T are Japan’s, $10T are China’s & $30T are Europe and “everybody else”.

Using that pesky “rule of thumb” (it works in reverse on the way up)… if Central Banks increase Interest rates, for every 1% global rate hike, the world’s bonds decline about 9% in value……so the first 1% global rate hike will cause about an $8 Trillion loss of wealth…..roughly the same impact for the 2nd hike.  So if rates increase 2% over the next few years, the total bonds outstanding will lose about $16T in value give or take (longer terms get hit harder than shorter terms) …..this mathematically predictable wealth loss is just shy of annual US GDP….Ouch!.  Does this sound about right?

No wonder Central Bankers are in no rush to raise rates…..and in fact, I have begun to believe they can’t without a carefully choreographed consensus.  I discuss this a bit in my “Theory of Financial Relativity” post…..  Any thoughts?   Perhaps in the context of Central Bankers losing control of Interest rates?  Currency relationships?  Disrupted exchange rates?  Will the markets take it upon themselves to “Gap-Up” until enough liquidity is (again) dumped into the global financial system?

Tim Bergin:

To back up for a second, the lack of liquidity in corporate bond markets is not as nefarious as it sounds.  The truth is that the Volcker rule and capital regulation has changed and that has led to less liquidity.  Despite having been a proprietary trader at a bank for over a decade, I have come to agree with the logic that banks should not be in the trading business. 

The reason is that Banks, like other 'AAA' rated credits (think GE – and the needed bailout of GE Capital), have too much incentive to increase leverage to capture small market spreads that their funding advantage allows them to earn.   For example, a popular trade for banks was to buy illiquid corporate bonds, buy credit default swaps to hedge credit risk, and then hedge out interest rate risk with interest rate swaps.  This complicated set of trades would allow the bank to make ~0.25% annually ( 0.25% is pre-crisis, more like 0.5% to 0.6% now, was as wide as 10% during the crisis).  Doesn't sound like much, but if you lever that trade 100x (something that capital rules used to allow), then it become much more interesting.  The risk is that you own a portfolio of very illiquid bonds, and you are subject to funding risk.  This trade lost banks many, many billions in 2008/2009

This activity, in the good times, is too easy for banks to pass up.  It is much easier money, and requires many fewer employees, than hiring people to lend to actual businesses.  The reality is that capital markets are an 'easier' way to make money; less people are needed, you can scale it quickly, and there is an illusion of liquidity to owning such assets.  This, I believe, has contributed to the asset market inflation.  This reliance on using financial markets to make money, rather than actual work, is the only thing I would add to your brilliant description of current markets in your  "Theory of Financial Relativity” post.

What changed was the Volcker rule came and said that banks can only facilitate client trades, not trade for its own account.  While well intentioned, the rule has made a mess of things.  One thing to note is that the bond markets are still largely traded by traders calling each other, not on electronic screens.  Each trade is usually a negotiation where the market maker gets to decide if he wants to trade the market indicated, with the counterparty he/she is engaged with.  So if you are a market maker and are in such a negotiation and buy a bond, unless someone has a mind reading device, it is impossible to know if that trade was proprietary or not.  Does that trader know he has another client that wants those exact bonds?  Did he just buy because they were 'cheap'?  Who knows.  As a result the market makers will try to satisfy their big clients, Blackrock, Pimco etc, by actually trading on the markets they indicate.  The rest of the clients aren't afforded that same level of service, as a result of the Volcker complications.  The market makers aren't doing anything wrong, just responding to the regulations.  To give you a sense of how complicated these regulations are, there are 1,000s of pages trying to describe what is market making versus proprietary trading.

I personally believe they should have kept it simple, if you don't want banks trading for their own account just repeal Glass-Steagall.

Not all the regulations were bad.  Capital rule changes have reduced leverage at these same banks.  So some of the excess leverage in trading books has greatly diminished.  This is especially true with derivatives.  It has created perverse and strange outcomes.  For example, swap spreads (the difference between the yield of interest rate swaps and the yield on US treasuries) are negative.  This was unthinkable, and would have been considered an arbitrage opportunity in different markets. 

Now that this is cleared up, on to your other questions.

The quick answer would be one of my favorite quotes 'great minds think alike, and fools never differ'.  I had just written a piece to my clients that looked at similarly gigantic numbers and came to very similar conclusions.

To quote directly from my last report (I have sent you a full copy):

In previous crisis periods, when equity markets fell, bond markets rallied as central banks lowered rates.  These inversely correlated movements provided a buffer for many portfolios. Losses felt in equity holdings would be offset or mitigated by gains in bond holdings. In the scenario I foresee, rising interest rates not only create large losses in fixed income markets, but also cause equity markets, whose valuations have been predicated on low interest rates, to also experience losses.

How bad could the losses be if both fixed income and equity markets fall?  Total global equity market value is approximately $80tril, G20 total non-financial debt (meaning corporate, and all levels of government debt) to GDP is 220% which means total debt is roughly $246tril.  If we assume interest rate markets correct the full 300bps I mentioned above, and assume the average duration of this debt is 5 years.  Losses in bond markets would be $37tril!  On the equity side, not all stock markets are as overvalued as the US market, I think at a minimum a 20% bear market correction could occur.  That would mean equity losses of about $16tril.  While $53tril in losses seems too high to be true, in this scenario losses would likely be higher.  Real estate prices would also drop with higher mortgage rates.  This would impact the creditworthiness of many loans, as property is a main source of collateral.  Taking into account credit and real estate losses, $53tril is likely too low.

To put these losses in context, the estimated losses during the financial crisis was $22tril.

While these losses seem huge, I am not suggesting a financial collapse.  Most of the losses in global bond markets would be mark-to-market losses, ultimately the majority of these bonds would mature at par, and investors would eventually be made whole. However, these mark-to-market losses would wipe out many overlevered financial market participants.  Depending on the nature of the participant, this could create some disorderly market activity.

I think most of the losses would be felt by non-bank participants, and so it shouldn’t be as big a deal as it was in the financial crisis.  Sure the bond holdings of banks and insurance companies would suffer losses but given the capitalization levels of banks (primarily US banks) I don’t think this would be a systemic event for them.

That said, countries that have experienced large property market rallies in the last 10 years would see bank stress.  The correlation between property market price drops and banking crisis is one.  So countries that I would be concerned about would be the likes of Australia, Sweden, UK, and of course Canada.

Will we experience a market with raising interest rates and dropping equity valuations?  I don’t know but the longer current conditions persist the more likely such a correction is.  The main point of this article is to highlight just how dependent the world is on low interest rates, and the danger this dependence poses to markets if such conditions change.

We have different amounts of total debt.  I think mine includes corporate debt, municipal, etc. and I am using a 5yr duration for calculating my losses.

I don't think it much matters getting the exact accuracy correct when looking at such calculations.  You just have to know that the world has not seen a market with declining bond and stock markets.  Losses in that scenario will be huge.

As you describe in your previously mentioned blog post it is a paradox.  What banker wants to be the one that triggers such losses?  However, can we permanently stay in an environment of zero interest rates and asset market inflation? 

My thinking has been that economic conditions are really starting to suggest that central bankers may have their hands forced. 

If we look broader and examine the economic data of the following group of countries; Canada, USA, EU, UK, Japan, and Germany.  The average GDP is 2.55%, unemployment is 4.9%, CPI is 1.5%.  Those are decent economic numbers, dare I even say 'normal' economic conditions.  All of these countries are at what is considered 'full-employment'.  Theory would suggest that inflation should increase under such conditions. 

Now contrast the economic data with central bank set overnight rates that average 0.33% for the same group.  Does that not seem far too low?  If that wasn't bad enough, the same countries are engaging in $109bil of QE per month!  Although $70bil of that amount should end by the end of the year.

Crazy stuff.  I think it may be even crazier to suggest that this can be sustainable, and remain under the control of the central bankers.

To highlight this in simple investing terms; German CPI was reported at 1.7% in December.  What idiot would buy 10yr German government bonds that yield 0.5%?  Yet that is where they trade.

Deep-Throat-IPO (Me):

Very thought provoking Tim.  First, I'm trying to recall, and I'm at a loss, as to any point in financial history when both bond values (Interest Rates Gap-up) and equity values fell simultaneously.  I could be wrong, but, as you say, I don't believe it's ever happened.  Perhaps the early 1980's with the Continental Illinois failure and the collapse of Mexico LDC debt?  If any of my readers want to chime in here I'd be grateful. The "double whammy" scenario seems plausible but I normally don't invest in scenarios where there is no precedent.  On the other hand, the world has never been in a global, seven year, near-ZIRP environment, where the "second largest economy" is based primarily on financial fraud and a fake dual-currency.  I guess we're in a brave new world.

Second, regarding the disparity in our "debt" figures (your $246T vs. my $91T) I would imagine that the difference is un-securitized debt.  My figure represents Bonds (Debt Securities) per the Bureau of International Settlements (BIS) your figure includes, by definition, all sorts of un-securitized debt (consumer debt, credit cards, shadow debt, etc.).  This un-securitized debt wouldn't necessarily be marked-to-market, but in the scenario we're describing, undoubtedly much of it would become non-performing overnight.  In short, we're on the same page.

The only other number that I'd feel comfortable updating is the global equity number.  The number I have was  just under $100T as of the beginning of December.  Based on what's happened with the markets in the last month, the number is probably just a bit north of $100T now, with roughly $35T of the run-up occurring in 2017. In any case, we are in agreement, when/if the "double-whammy" reset we're discussing takes place, the losses will be enormous.

The only other tidbit I'd share is that the valuations of US equities, and presumably other "less regulated" (I'm being kind here) global markets have never been higher.  I'd suggest that, based on current valuations, your estimate of a 20% bear market, if/when central bankers lose control of this mess might be relatively optimistic. 

To wit, I completed a valuation study in April of 2017 in a post entitled "An IPO Investors Snapped Up!"  describing the silliness, fraud and froth out there in the US equity markets.  The result is that only the top businesses by sector had any earnings to speak of. 

As the above chart describes, the "Top 50" Companies in each sector currently provide 88% of all earnings for publicly traded US Companies.  The remaining 4,492 businesses, to be blunt, are in most cases a mis-deployment (waste) of capital.  They are all valued as though they are the next Apple when in reality nearly all of these businesses, over time, will prove to be "dog shit".  Money has to go somewhere.... unfortunately here's where it's choosing to go.

I'll probably take the time to update the study in a month or so, just to see where we're at, but my guess is that it's even worse today. 

Moreover, the Shiller CAPE is as high as it's ever been (with the exception of the dot.com bubble) at 33.5.  The 500 largest, best-managed, publicly traded US businesses are priced at 33.5 years worth of "average" earnings.

When you factor in all of the above, plus the probable financial chicanery of the Chinese Real Estate & Bond Bubbles as well as the fake ADRs listed on US Exchanges, my guess is that if (and only if) Central Bankers lose control of this, (i.e. the Theory of Financial Relativity is dead wrong) we would gladly settle this revaluation for $53 Trillion, take the loss.... and move on.

Our only hope is that the TOFR holds true and Central Bankers can indeed join hands, sing Kumbayah and somehow keep all of the balls in the air.....forever....or at least as long as they can until they form a consensus  and ever so gently, nudge interest rates back up.   If our Central Bankers succeed, anyone fortunate enough to hold financial assets will get rich while the rest of the world somehow tries to survive on a minimum wage.  If the Central Bankers fail, average folks will wake up one day to find that their Rip Van Winkle retirement plans have been wiped out.

Finally, to take a stab at answering your rhetorical question: "What idiot would buy a German Bond?"  Wilhelm Rogers would have!.....he was less concerned about the return "on" his money than he was about the return "of" his money.  In other words, if your/our scenario is right, massive losses unwinding this mess, a return of 0.5% will look pretty "gute"....."Ach du Lieber!

Anyway Tim, thanks so much for taking the time to discuss these interesting and complex topics.  It was absolute pleasure.

All the best in 2018!

India Pictures

One last thing.....I don't think I mentioned that I'm writing from India right now.  As always, we're enjoying our trip.  My wife suggested that, instead of posting the usual "goofy" pictures I post from our trip(s) that I should post a few of the more beautiful sites of India.  I think that's a wonderful idea and like all good husbands, I'm happy to oblige.  Below are a couple of pictures of a local Ashram and the garden at our family home.  The flowers are beautiful this time of year.

For all my good friends back in Cleveland, I understand that the temperature is below zero right now.  On the other hand, for my Canadian friends, those temps are just a nice spring day. 

As always, it's all about perspective.  Be strong.....