Reading the excellent article by Brian Meehan of Bloomberg Intelligence, “Basis Trade Growth Is Massive”, I was reminded just how stupid clearinghouses really are. If you ever meet a head of a clearinghouse, I can assure it was not brains that got them to this key position in global finance. As the article points out, their is now a USD 1.2 trillion notional short position in US treasury futures. You should not read this as the market being bearish on treasuries. It is a levered trade to make a “risk free” return on the difference in price between off the run treasuries and treasury future positions. For every short position in the treasury futures, there should be a long position in the physical market.

The size of the trade has been large for awhile. The Federal Reserve took a look at it and thought it was not a problem (uh oh!). It has the same data as above, just that Bloomberg has provided an up to date estimate. In the post GFC world, much of this trade has migrated from prop desks at banks to hedge funds.

This is not a new phenomenon. BIS noted in 2019 reports that the Repo market has become bifurcated between banks and hedge funds. See dotted line in middle graph.

This is a natural by-product of how clearinghouses work. The clearinghouses at the centre of repo trading (LCH and CME) are only concerned with how quickly trades can be liquidated. As they tend to settle on a daily basis, and hedge funds and other asset managers need to settle constantly, they tend to allow trades to grow in size much more than when investment banks were at the centre of repo market.

Clearinghouses cannot go bust. So they pursue policies to grow volumes as much as possible. One such policy is compression trades. This is where very similar interest rates products can be compressed into a single product to allow more leverage to be used. Below show that a multiple of outstanding notional is being compressed every year.

The problem is that when there is a shock in interest rate markets, then liquidity disappears and clearinghouse starts demanding more margin from all players, which then leads to spiralling weaker market. We saw just such a circumstance in the UK gilt market under the PM Truss. The sell off in gilts could only be contained by the Bank of England becoming a buyer of last resort.

The big investment banks know that the clearinghouse misprice risk, which is why they suggested that they should take more financial risk in pricing, to incentivise them to do a better job. Regulators have so far ignored this advice. The proposed risk waterfall is presented below. More detail here. A close look at the proposed margin from the waterfall is a the ability to bankrupt CCPs. This has so far been rejected.

So how does this fit in with LTCM? Well LTCM focused on risk free arbitrage trades. The problem with most of the trades is that you need substantial leverage to make them work. Most banks would feel uncomfortable lending a hedge fund that much. The downfall of LTCM as it was told to me, and I did work at UBS, was that one day there were industry drinks and one trader mentioned how much business he was doing with LTCM - and another trader said, hang on a second, we are doing just as much business as well. The traders then worked out that LTCM had far more leverage than they had disclosed, and as they traders worked this out, sold their positions before LTCM could. This put downward pressure on their trades, eventually forcing them bankrupt and bringing about a bailout. In essence, LTCM became the market, and at that point, it was destroyed. But it only became so leveraged because it acted in secrecy.

Under the new clearinghouse system, no such subterfuge is necessary. The clearinghouse actively wants you to borrow as much as possible, as long as you meet the liquidity and volatility seen over the look back period - typically 10 years. The problem, as the details above show, is that the hedge funds are all in. So are retail investors. Shares outstanding in TLT has skyrocketed.

So what could go wrong? Well in the US, after the presidential election is out of the way, you could expect inflationary policies to be announced (higher minimum wage, more tariffs, more stimulus). This leads to inflation expectations rising. Physical treasuries are held in ETFs like TLT, which probably see redemptions. Treasuries fall, basis trade widen, margin calls are issued, which leads to more selling of treasuries, and more basis widening, until the Fed steps into buy treasuries. This seems likely to me. My long GLD/TLT trade should do well in that environment. Having gone sideways for a few months, a move higher looks overdue.

The clearinghouse reform is failed reform - but it will take a crisis to change. Maybe this will be the one.

Subscribe now

For a long time I have said the BOJ was the most powerful central bank in the world, and we have prove of it today. Raising interest rates, and reducing QE, has turned Nikkei from an up trending market to bear market very quickly.

A collapsing Nikkei tends to be seen as a indicator of incipient economic doom. But even with the moves today, the Japanese 30 year still seems to be in an uptrend.

As is the relative Japanese bank trade.

The problem with the deflation story is that it that bond yields have already proven themselves to be a policy story more than a macro story. Do you remember when every economist and hedge fund manager was telling you that the unwinding of the Chinese property bubble would be the GFC on steroids. Chinese property developers have gone bust in droves, and property starts in China has collapsed. And yet US bond yields have remained fairly high, and JGBs remain in bear market.

Looking at US policy going forwards, both Trump and Harris I think would like to restore US manufacturing to its former glory. Using non-farm payroll data, we can see how much more employment would need to rise in this sector. The US lost 6mn manufacturing jobs from 2000. Could policy restore this? I think they are going to try - and it will not be deflationary!

Curiously, the number of US government employees has recently risen to new highs, after a long period of stagnation.

Hopefully this helps you understand where I am coming from. Deflation had little to do with economics, and more to do with government policy. Inflation has little to do with economics and more to do with government policy. For me, policy is inflationary. So why have markets moved so violently? Well I have to go back to the way clearinghouses price risk - which is backward looking. So even though the BOJ has been talking about raising rates, clearinghouses not have anticipated any volatility until it actually happens. Once volatility spikes, THEN they ask for more initial margin, which then causes more volatility and so on so forth. As I pointed out, clearinghouses create more financial volatility, and likely bankruptcy. Their only “benefit” is that governments can bail out market via clearinghouses rather than banks. The Nikkei VIX chart give you a good idea of how bad they are at pricing risk. Nikkei vol trended higher into the GFC, but todays 70 print came from nowhere. Margin call is the answer.

I ask myself, is the political will there to create unemployment to defeat inflation? I don’t think so. The BOJ is reacting to this change very slowly, and leveraged players are going to go to the wall, but I don’t think we are going to see austerity anytime soon. The stupidness of clearinghouses is going to let you get in on inflation trades at good levels I think.

GLD/TLT still looks good to me.

Subscribe now

In 2022, clearinghousse were at the centre of two large capital destructive moves. LME was a the heart of the spike in nickel price, which was resolved by LME cancelling trades to the benefit of JP Morgan and its clients. We then saw the collapse of the gilt market, driven by margin calls by LCH and could only be resolved by the Bank of England stepping into the market. I have long argued that the clearinghouse model is more prone to crisis and panics than the bank centre model that it replaced. I have taken a closer look at LCH after a couple of subscribers sent me this news release from LCH.

In an interview the head of this new service estimated that initial margins could fall by 50% or more (link). The problem is that collapsing initial margins can then lead to a crisis as the clearinghouse then has to raise margins quickly in response to a crisis. Perhaps the most surprising feature of this announcement is that this is been carried out, while the issue following the collapse in gilt prices in 2022 have yet to fully play out at LCH. So I have gone through the IOSCO- CPMI disclosures in detail to see what we can learn.

The first thing it should be pointed out is that LCH and CPMI-IOSCO disclosures are not particularly transparent to outsiders. Data is disclosed on 8 separate default funds (the mutual pot into which all members pay) - they are Equities, Fixed Income, OTC_FX and Interest Rates for the UK based Limited part of LCH. For the European based LCH SA - the default funds are Cash and Derivative, Fixed Income, OTC CDS, and €GC Plus. Each default fund seems to match up to a clearing service, but the names are not a perfect match which is why I say seems. For example, the Equity default fund would seem to match up EquityClear clearing services, while the fixed income default fund seems to match up to RepoClear. And €GC Plus seems to match up the Tri-Party Repo Clearing Services. LCH discloses data for each default fund and clearing service each quarter. There are over 200 data points which are disclosed each quarter, and LCH produces an excel spreadsheet with 12 tabs.

Having looked at IOSCO-CPMI disclosures for a few years now, and having seen the meltdown in the gilt market, we can try and pinpoint the important data, and try understand the risks in LCH. First of all we can look at peak breaches by clearing services. Equity breaches were tiny, so left out of this graph. For those of you at home, this is disclosure 6.5.4. Note this is peak breach in last 12 months, so when it plateaus, it means that most recent quarter did not see a higher breach (I have found the numbers to be occasionally revised without warning). As you would expect, RepoClear and Interest Rates saw big breaches, of a total 1.6bn GBP in Q3 2022.

So for the Limited, UK based clearinghouse, we know the Repo and Interest Rate clearing was where the problems were centred on. Oddly, disclosure 6.5.1.1 gives us the number of times margin was breached in the quarter. So even though the size of the margin breach in the clearing service was similar, there were peak of 237 margin breaches for Repoclear and nearly 19,000 for Interest Rate clear in Q1 2023. What I do not understand is why margin breaches continue to rise even as interest rate markets have settled down.

Peak breach amounts would point to RepoClear as being as large a source of risk to the clearinghouse at its Interest Rate business. Oddly, the maximum variation margin paid to the clearinghouse for these businesses would suggest RepoClear to be too small to be that risky.

The best measure of risk actually comes from initial margin. This is the amount paid by members of the clearinghouse to ensure they get paid. The volatility in the interest rates initial margin, where it has twice climbed from GBP 1bn initial margin to over GBP 5bn in the space of a few quarters.

What do these data points say to me? The interest rate business is where the majority of volume is traded at LCH, but the relatively low initial margin to variation margin suggests that this volume is concentrated in high volume low duration assets, with little risk of capital loss. This makes sense to me. The equity business seems to be correctly priced, with high initial margin, and little margin breach. The standout risky business seems to be in the Fixed Income/Repo business. A simple risk measure to look at would be size of peak margin breach to initial margin. As initial margin is contributed by both sides, when a breach is approaching 50%, it means the losing side as burned through all their initial margin.

Why is the fixed income peak breach to initial margin look so different to interest rates in most recent quarter? For some reason, initial margin fell at the fixed income/repo even though it doubled initial margin double interest rate. Very strange. One of my big problems with clearinghouse is that they are incentivised to get initial margins as low as possible. The fixed income/repo market at LCH also distinguishes itself by having a different membership make up to the other default funds.

To recap from my last post on clearinghouses and LDI, LCH allowed a number of asset manager to directly access the RepoClear directly. Here are all LCH members that are only members of RepoClear.

Most notably is the inclusion of a number of LDI funds. Sadly, these funds are no long listed on Bloomberg, but I can repost the performance from my last post on clearinghouses.

The profit motive of clearinghouse leads them to seek to increase volumes, and increase membership. What we have seen with LDI meltdown, nickel market, and European power trading before that, is that the system is guaranteed to become unstable. The good news for regulators is that when it threatens a large bank like JP Morgan in the nickel market, the clearinghouse can organise an indirect bail out by cancelling trades, or the BOE can intervene in the market to provide liquidity, without having to directly bailout large banks. Small investors without political power can be left to go bust (see Silicon Valley Bank for another example of this regulatory bias to large corporations). The clearinghouse model remains for me a failed financial reform, but as a process to allow back door bailout of large banks, it remains unmatched.

Subscribe now

I was reminded to look at clearinghouse by the beginning of lawsuits about the nickel debacle at the beginning of 2022. The fact is that nickel is a bit of a niche market, the really important market is in interest rate products. Back in September, the Liz Truss budget cause a sharp sell off in UK bonds (gilts). This in turn caused problems at LDI funds in the UK who were hit with margin calls, and prompted the Bank of England to purchase gilts to stabilise the market.

The above graph is taken from a Bank of England report which blames the funds - particularly LDI funds. The real culprit is the clearinghouse, but as they were following regulator’s directions, I guess the BOE feel that they cannot be blamed. For long time readers, I have highlighted that the move to clearinghouses at the centre of the financial system has changed the market in a very fundamental way. The first thing you need to know is that the clearinghouses, with the regulators full support, have wanted more asset managers to be members of clearinghouses. They have largely succeeded as can be seen below. Please note the increase in non-banks in clearing in right hand side graph.

If you go to the LCH website you can get a full list of members. While I expected to see asset managers (usually large pension funds) to be listed as members, I was surprised when I came across LDI Funds as members of Repo Clear. As you can see below, there are 13 LDI funds. As far as I can tell, they are all managed by Insight Investment Management Group a £653bn asset manager. Insight is owned by Bank of New York Mellon, which is also a member of LCH - although not in RepoClear. Why do funds owned by a member of the LCH need to become direct members themselves? My assumption is that perhaps Bank New York Mellon did not want to either take on the trades that the LDI funds were doing, or be responsible for the margin calls (same thing really).

The last two funds listed above in the LDI section - LDI Enhanced Selection Longer Real Fund and LDI Solutions Plus Partially Funded Index-Linked Gilt Fund are both listed. Performance is shown below.

Without knowing the specifics of why LDI funds were direct members of the LCH I can only speculate on the reasons for this, but to my mind it stinks. The reality is that the banks used to regulate the repo market themselves, with the majority of trades between banks themselves. As can be seen in middle graph below, banks used to be borrowers from the repo market, but as banks have increase their holding of treasuries, in 2018 the US banking sector became net lenders in the repo market. The implication is that asset managers have to be the borrowers. So on one side you have large sophisticated banks with a long history of operating and pricing repo, while on the other side you have asset managers. The clearinghouse is profit driven, so is incentivised to get capital costs as low as possible. If the asset manager is passively run (that is the counterparties know exactly how they will be positioned), then the chances that the passive manager is going to be on the wrong side of a trade is going to be very very likely.

The final problem, and what causes markets to be volatile and suddenly destroy capital is that the clearinghouses have gone out of the way to reduce the amount of capital in the system. This has been done in a number of ways. Most transparently is the trend to have low initial margin.

Low levels of initial margin mean that margin breaches are then margin calls become more likely.

LDI funds have been granted direct access to clearinghouses to try and increase liquidity. The clearinghouses set risk based on historic volatility, and raise margin as volatility increase making markets MORE volatile. This has now happened at least three times - the reop blow up in 2019, the nickel blow of up of 2022, and the LDI/Gilt blow up of 2022. I could potentially add treasury market illiquidity in 2020. The clearinghouse at the centre of the financial system is failed reform in my mind. But the regulators keeps blaming funds. This is even after the banks have told the regulators that the clearinghouses are badly run (See “A Path Forward For CCP Resilience, Recovery and Resolution) As long as cleringhouses are profit seeking and remain at the centre of financial market, expect further market blow ups, and expect asset managers and retails clients to continue to lose money. A regulatory change that makes a financial system more unstable, while systemically being disadvantageous to retail clients is not a regulatory improvement, its embarrassment and a scandal.

Subscribe now

I am going to use some technical details from other clearinghouse presentations that I have done, so if you want more details click on the clearinghouse tab on the home page. Some of them are behind the paywall, so you might need to become a paid subscriber or start a free trial.

I personally do not think LDI investing, and pension funds are to blame. In a world where Japan still has negative interest rates do you really think there not the fund available to buy 30-year gilts at 5% yield? Or even domestic UK institutions that would be happy to look in sizable capital gains? No the answer lies in chronic risk mismanagement at the LCH. LCH is the largest interest rate derivative clearinghouse in the world. It dominates European interest rate derivatives and has a sizable position in US markets as well. So how did LSE owned LCH ruin the bond market? Just a quick recap. Prior to around 2011 or so most interest rate derivatives were traded bank to bank. But regulatory reform forced banks to push all trades through a clearinghouse.

As more and more trades became centrally cleared, LCH became aware of more and more circular trades. That is Bank A might have a trade with bank B, and Bank B would have an identical trade with Bank C, and Bank C had an identical trade with Bank A. LCH would then call up the banks and say hey guys, you don’t need these trades anymore, we can just cancel them and return the risk capital we held against the trades to you. This is a great idea in principle, and the banks and regulators all engaged in this. These trades are called compression trades, and LCH was happy to report that it has grown this business from 200 trillion USD to near 900 trillion USD in 2019. This compares to a notional outstanding of USD 400 trillion USD. Note the physical gilt market is only 1.5 trillion USD, so can get how derivatives can dominate the physical market here.

If you talk to LCH or hedge funds or traders, they all love compression. It simplifies the market and frees up capital and initial margin. But if you step back and have a think about it, you can see where the problem comes from. When a pension fund or a hedge fund agrees to a compression trade, is that fund actually reducing the bets it has on the market? Are they actually reducing the capital at risk of their position? No, they are not. But they are reducing the amount of capital is the system. So if risk is not being reduced, but the amount of capital in the system is being reduced, is the system becoming more or less stable? Of course not. These are simple questions that somehow the regulator forgot to ask. The first problem with clearinghouses is that they basically removed risk capital from the system.

The second problem is that clearinghouse only price risk on past performance, and with no reference to value. Most clearinghouse price risk of a 10 to 7 year look back. Which means that the risk (or capital that traders need to trade gilts) coming into to 2022 would be quite low, as the gilt yield had fallen in a slow and predictable manner for the past 10 years.

So why does the BOE have to buy gilts? So clearinghouse being braindead, idiotic machines now look at the price history of gilts and says “These are no longer safe assets - everyone needs to place huge margin to trade these things” When in reality, at 5% yield they are probably closely to being correctly priced. That is clearinghouses are now physically stopping the financial system from buying bonds. Of all the post GFC reforms, placing clearinghouses at the centre of the financial system has to be the dumbest. End Rant.

Subscribe now

Clearinghouses were meant to reduce risk in the financial system, and reduce the risk of bailouts. In practice, they make market bailouts compulsory. The most recent example is the UK gilt market. As a recap, the UK announced a budget that both increased spending and cut taxes. Markets reacted by dumping the pound, and selling bonds. Investors in long dated gilts have lost well over half their value this year as yields have risen from 1% to nearly 4%.

So far, so standard. Fiscal recklessness causes investors to dump the currency and bonds, and the typical reaction here should be a tightening by the Bank of England to calm nerves. And financial market have reflected that reality in pushing up the yield on 2 year gilt to over 4%.

But then the BOE did something very odd. They restarted QE. That is they announced a program to buy bonds - which is complete at odds with a central bank that is trying to combat inflation. In the end, they did not need to conduct much QE - but they were prompted by fears of “dysfunction” in the long dated (over 20 years) gilt market. So how is it that the long dated gilt market could suffer dysfunction? The same way that the nickel market suffered dysfunction, and numerous other market in recent years - clearinghouses.

Read more

Subscribe now

As picked up by Robin Wigglesworth at FT Alphaville, @mechmarkets on Twitter highlighted that Chris Edmonds, the Global Head of Clearing & Risk at Intercontinental Exchange (ICE), was in a position to put a major financial institution into default in March 2020, but chose not to. The CFTC have the YouTube video up, and you can find the link via @mechmarkets . So can we use ICE regulatory disclosure to shed anymore light on this event? ICE manage seven different clearinghouses. As a general rule, the more initial margin at a clearinghouse, the more important it is. Using ICE data, only the ICE Clear Credit (ICC), ICE Clear Europe (ICEU CDS and Futures & Options) and ICE Clear US would be big enough to be a problem.

Financial trouble would be caused by the clearinghouse two ways. Firstly, asking for a large increase in initial margins. If this was the case, then ICC will be the place to look. It had by far the largest increase in initial margin in Q1 2020.

The other way you can get a big problem is that if the market is bifurcated, and requires huge payments from one set of investors to another. This creates outsized variation margin flows. In this case, we should look at ICUO F&O or ICUS.

What stands out to me is the very large increase in initial margin relative to the change in variation margin that occurred at ICC. A big shift in initial margin occurs when a good credit it suddenly perceived as bad. All the other ICE clearinghouses saw increases in initial margin as levels lower than the increase in variation margin. Focusing on ICC makes life much easier to work out who would be in trouble. ICC members are made up exclusively of large banks, and as Chris Edmonds states the potential defaulters is in the room, narrows the suspects down to Citi, Goldman Sachs, JP Morgan and Morgan Stanley. ICC seems very likely to be the right place to look, as the Federal Reserves intervention into credit markets allowed this market to rally quickly, and would have allowed any bank that was facing a huge margin call, to get out of trouble. The CDS of these four main banks show exactly how quick the reversal was in 2020. I should point out that which bank was in trouble is a mute point - if any of these banks defaulted, they all would have been in trouble.

What LME and now this disclosure from ICE highlight, is that the clearinghouses will and do bend the rules for their large clients, the banks. The clearinghouse system has mutualised risk, while at the same time placing glorified civil servants at the centre of market pricing. The stated aim of the clearinghouse reform was to make sure no bank was “too big to fail". The end result is that clearinghouses ensure with ad hoc rule changes, “no bank can fail”. Perhaps the real aim of regulatory reform is to remove the power of banks to force another bank into default.

Subscribe now

As mentioned in previous posts, ISDA provide a Standard Initial Margin Model (SIMM), is reactive to price movements, rather than forward looking. In my understanding, the problems in clearinghouses are most likely to occur when a trader is doing a spread trade between two very similar, but not identical products. So in my thinking, problems are likely to occur where clearinghouses believe similar products are very highly correlated. One area this could happen is in interest rate products. Clearinghouse consider the correlation between a 20 year bond, and a 30 year bond to be 99% for example.

The other thing you need to look at is the “delta concentration risk”, which basically sets an upper limit to how much a position can make or lose. The ISDA-SIMM details concentration risk for credit spread trades, which seem to me to rather low, with the exception of mortgage backed securities.

Within the equity space, clearinghouse risk pricing is a bit harder to follow. To summarise, a stocks will be classed according to its size (large cap, which is defined as a market cap of over USD2bn, or small cap), geography (emerging or developed), and finally 4 separate industry categories. Generally speaking the higher the risk, the lower the correlation. However, there was one area that caught my eye. Indexes, ETF, and Funds (bucket 11) are considered the least risky, no matter what their composition. This would mean that for investors interested in holding a risky asset, from a capital perspective, you should always hold it via an ETF.

Within the equity space, the one area that has a high delta concentration threshold is ETF and Index based products. This would imply that a fund can leverage up holdings of ETFs far more than single stock holdings. And in fact would prefer an ETF to a stock. ETFs also have the highest correlation discounts, so I would expect problems to come from ETF liquidity issues if clearinghouses were a problem. Very low liquidity if S&P 500 futures may be a sign of this.

However, the one stand out area for high correlation discounts and large delta concentration thresholds is energy and particularly oil. So two different crude products, such as European and US oil would be given 98% correlation.

What we also see is that between different crude oil buckets correlations are very high. So diesel prices soaring above crude could also cause problems.

Delta concentration is also very high for crude oil and natural gas. Note nickel is a base metal, with low correlation offset and a low delta concentration threshold, and yet still nearly blew up LME.

Looking at the example of Einar Aas, and his power trades, my guess would be that large moves in oil spread trades would be most likely to cause problems with clearinghouses. If we look at the June/July 2022 spread for Brent, we see the sort of spike that should cause problems.

Despite this move, the clearinghouses have shown no sign of distress. We have seen some distress in oil trader Trafigura’s bonds. But these are already showing signs of recovery.

The lesson I take from this is that clearinghouses can keep the financial system from suffering solvency issues, as long as liquidity is being provided. The question that remains to be answered, is when liquidity is withdrawn, does this turn liquidity problems into solvency problems?

Subscribe now

One of the big problems with financial markets is survivor bias. A good example is the bond market, where investors who survived the bond rout of the 1970s would constantly caution younger investors on the bond market throughout the entire 40 year bull market in bonds. Shorting JGBs was such a terrible trade, it earned the moniker “widowmaker”. I have wondered for a while if US equities, and particularly the S&P 500, was becoming something similar to a JGB trade. I am starting to think the clearinghouses provide a way for us to think about the S&P 500 as a new JGB trade. First of all, using historical analysis would suggest a number of reasons to be bearish on US equities. Typically, when inflation is rising, the S&P 500 has struggled. 1970s saw rising inflation, and poor returns from the S&P, and likewise, the very mildly rising rising inflation of 2000s, also saw stagnant returns.

But the more I think about markets and politics, the more I realise that governments react to what societies want. After the inflationary 1970s, policy making was set to combatting inflation. If that meant unemployment and bankruptcy rose, then so be it. Hence the natural tendency of markets to underperform as inflation rose. But the structural changes made by politicians meant that inflation was always going to surprise to the downside. So to continue the analysis, after the GFC and then the Eurocrisis, voters and politicians basically did not want financial crises anymore, particularly ones that led to soaring unemployment, and this has led to institutions intervening in markets to ensure stability. So in my last “Buy the Dip or Sell the Rip” note, I was looking at HYG US, and saying it was still saying sell the rip. The yield on high yield is also still far below CPI.

When I look at the VIX curve, Spot VIX is again trading at a big discount to UX8 (8mth generic VIX) which is a sign of complacency.

However my recent work on clearinghouses had made me think about markets with fresh eyes. The clearinghouses do not set risk with an eye to future inflation. They also don’t set risk prices with an eye on how leveraged underlying participants are, or even how they may all have the same positions. The clearinghouse is only concerned with liquidity. The less liquid, the more risky the asset is for clearinghouse. Which is probably why large cap US stocks have been able to trade so well for so long. Now, as we seen with the LME, and Nasdaq Oslo, one trader can destroy those clearinghouses, but unlike the LME and Nasdaq Oslo, western credit markets have been able to rely on a very large player to provide all the liquidity they need to keep asset markets moving - central banks. For that reason, the biggest and best “buy the dip” reason I can come up with is that the Federal Reserve balance sheet is still growing.

Market carnage was far more widespread in Q4 2018 when the Jerome Powell suggested that balance sheet reduction was on autopilot, that during Covid when the world went into a cardiac arrest. When I looked at Japan, as model for markets after the GFC, the markets would look through what the central bank and governments wanted, and prices assets correctly. Hence I was not a big believer in QE working. But I can now see that the replacing of investment banks with clearinghouses allows central banks to provide liquidity directly into markets, allowing asset prices to rise (as has been their policy goal since the GFC). The issue for investors is when do governments place reducing income inequality, or combatting inflation as their main policy aim? China has already moved down that route. The China dominated Asian High Yield ETF (traded in Singapore, but quoted in USD), shows what happens when governments withdraw support for credit markets.

So buy the dip, or the sell the rip is not a technical question of valuation or market positioning. It is a question of whether the Federal Reserve pulls liquidity or not. If central banks pull liquidity, then clearinghouses will reprice risk, regardless of the outlook, as they did in 2018, and in China in 2021 and 2022.

One reason I publish is to flush out different opinions or data sources. Or as one wag put it, if you want the correct answer to anything, just publish the wrong answer on the internet! So I am going to republish many of the charts I have used for clearinghouses, but with a user’s view, and then my comments.

Read more

Subscribe now

As there is a lot of interest in how clearinghouse work, let me explain in more detail how they price risk. The key point I have been trying to make in my series on clearinghouses is that by moving clearinghouses to the centre of the financial system they misprice risk. with the emphasis on pricing risk on a backward looking basis, rather than a forward looking basis. Bankruptcy tends to happen suddenly in a clearinghouse centre financial system.

To understand how clearinghouses price risk, we now have four overnight bankruptcies/market stresses we can look at. We are going to use the ISDA Standard Initial Margin Model (SIMM) for a proxy of how clearinghouses price risk before and after. We are also only going to look at two metrics from the ISDA SIMM - risk weight and correlation. With risk weight, lower is less risky. With correlation, higher is “better” as it means that hedging with similar products will allow you to reduce initial margin.

In chronological order, the first one we have in the bankruptcy and liquidation XIV US (the inverse VIX ETF). As a reminder, XIV was essentially bankrupted when VIX doubled overnight. Note that after the XIV blow up, volatility indices saw their correlation offset reduced, and their risk weight increased.

Einar Aas went bankrupt on wrong way spread trade in European power trade. Again note that after Einar Aas was bankrupted correlation for European power was reduced and risk weight was increased

We had the repo rate spike in 2019, which required the Federal Reserve to step into the market. Again we see that after the blow up, correlation offset is reduced and risk weight is increased. You could even wonder if the increase in risk weight in ISDA SIMM V2.2 cause the repo spike? The SIMM was published in early September 2019.

With nickel causing problem today, we will have to wait for the next ISDA SIMM update to see what happens to risk weights in base metals. Correlation will almost certainly reduce and risk weights reduced.

Off the four blows up we have had, all have been contained to some degree (so far). The question we need to ask is that are all four of these episodes tremors before a larger market move, or is this just how the new financial system works, and we need to learn to deal with overnight bankruptcies going forward? Part 2 will try and address this issue.

Subscribe now

I have been thinking about clearinghouses for well over four years now, and build a mental model for trying to understand how they get in trouble (see the early clearinghouse posts). Whenever I talk to someone about their actually experience with LCH, they tend to be very positive, and usually assure me it reduces risk, and is far better managed than LME. The problem is why did major banks, which are major customers and potential beneficiaries have such a problem with the clearinghouses? The clunk title below is basically code for the banks to say they want CCPs to hold more capital and to be put into bankruptcy if needed.

The author for Goldman Sachs is a Nicholas Friedman, who was involved in the margin call on AIG that began the GFC. If he is worried about clearinghouses, then I am worried about clearinghouses.

So I started asking myself, what do the banks see that the rest of the market not see? The clearinghouse offer copious information via CMPI IOSCO disclosures, but in such a way as to make it very hard to extract meaningful information. So far the best source of analysis on LCH I have found comes from the Reserve Bank of Australia’s Assessment of LCH SwapClear Service. One useful graph is the amount of initial margin held at LCH for OTC rate trades.

RBA also provide a useful graph on how initial margins have changed during Covid. With an initial spike dues to changes in market prices, which has fallen off, to be replaced with a some pro-cyclical charges.

None of the above graphs provide a good reason to understand why banks are worried about clearinghouses. However, in an assessment from 2020, we find a graph that may explain banks concerns better. Banks, which are direct participants, provide far less initial margin than clients (that is clients of members), which is in stark contrast to 2015. In essence, non members are bigger players in clearinghouses than members. As a bank, you are now forced to hope that the other members have done the correct due diligence on their clients, in particular their ability to meet margin calls.

As highlighted in previous posts, I believe more “speculative” clients, and aggressive compression makes me think that clearinghouses are becoming more risky, not less risky. One way risk is priced for clearinghouses is a measure called KCCP. This quantifies the amount of capital that must be held against default fund contributions. The lower this amount, the less capital that must be held. As you can see below, from 2015 until beginning of 2021, this measure was falling. The most recent disclosure has it rising.

A universal financial truth is that risk does not disappear, but it can be moved to those best placed to bear that risk. This is a positive feature of financial markets. But another universal truth, is that complicated financial systems can often obfuscate who is really taking the risk, which leads to mispricing and a boom and eventual bust. Clearinghouses look to have moved into the latter category for me.

Subscribe now

Subscribe now

Most investors now understand how the LME bailed out its member caught short nickel. And hopefully investors can now understand how placing clearinghouses at the centre of financial systems actually reduces transparency. With that understanding, I am going to revisit the Repo market blow up, and explain how this was caused by LCH. This blow up required the Federal Reserve to bail out a number of hedge funds caught on the wrong side of the trade by intervening in the repo market. My best guess is that the hedge funds that got bailed out have doubled down on this trade, and without the Federal Reserve restarting QE soon, should lead to chaos in the repo markets and the LCH. I have talked about the repo market before, but as a brief cap, the middle chart below is form BIS shows how repo markets have become bifurcated, with banks on one side and hedge fund on the other.

In my view, and strongly corroborated by the joint press release issued by JP Morgan, Goldman Sachs and other members of LCH, was that hedge funds had become extremely leverage on spread trades in the treasury market, and the repo market disfunction in 2019 was driven by these fund being on the cusp of blowing up. The Federal Reserve stepping into do USD60bn of repo a day, essentially bailed out these funds.

This leaves three questions in my mind. How did the hedge funds get enough leverage to make the treasury market dysfunctional? Who are the hedge funds involved? And finally are they still doing it?

Read more

Like many commodities, nickel price has surged from the Covid lows. However this week is has spiked, effectively doubling, and briefly reached highs earlier this week of USD 100,000.

In response to this move, the LME has closed the market for nickel, and cancelled trades. Why have they done this? Essentially this was to give time (and by cancelling trades, reduce the amount) for a major member to make a margin call. But in reality this was default by a member, and LME choosing not to place that member in default. If we look at the default by Einer Aas in 2018, we can see why this is a default in all but name. When Einer Aas could not make a marign call he was bankrupted. His default waterfall is shown below. His initial margin was taken, then his default fund contribution. Value unknown, but I would estimate around Eur 30m. The clearinghouse then chipped in Eur 7m, and then there as a Eur 107m from the commodity service default fund. This commodity service default fund is paid by “surviving” members of clearinghouse. As every clearinghouse trade is matched, typically “surviving” members are the winners, or traders that have been on the right side of the trade.

The loss on the LME Nickel trade has been estimated at up to USD 12bn. The LME default fund (4.2 on its IOSCO CPMI disclosure) is given at USD 1bn. I don’t know if that number is entirely accurate, of if all of it can be applied to the nickel market, or is segregated by product. Clearly, there was a counterparty who could not make margin call, and if the rules had been followed, they should have been declared bankrupt, and the position sold, and any losses would be paid by LME, and then surviving members.

However, LME has cancelled trades, to reduce the price from USD 100,000 to USD 50,000, closed the market until Friday, and will limit price movements to 10% a day.

What does this mean? We know that one major counterparty should have gone bust, and is now getting a stealth bailout from the clearinghouse. Rather than asking surviving members to recapitalise, it has artificially adjusted prices to reduce their winnings. All of this has been done without consulting all members, and to the benefit of as far as we know one well connected member. As mentioned before, in 2019 large clearinghouse members recommended reforming the default waterfall.

In this new model, clearinghouses would have to put up more capital, and clearinghouse members would choose whether a bailout would occur or not. Sadly, reform has not happened, and the surviving members of the LME will pay the price.

Initial margins are in financial terms very similar to option premiums. The more volatility in the market, the more margin there should be. Using BIS data we can track the notional interest rate derivative value outstanding, and the gross value of those derivatives, gives us something resembling implied volatility. As you would expect in a bull market, the premium value is only 2% of notional, and in a bear market rises up to 4 or 4.5%. Most recent observation is June 2021 at 1.8%. So this basic analysis suggest at least a doubling of initial margins on a cyclical basis.

The problem with this analysis is that it ignores the changes to the clearinghouse system. As pointed out in a previous post, compression trades now are larger volume at LCH alone than total IRD notional outstanding.

Read more

This note has been rushed out due to events in the LME and nickel today. Please listen to the video for views and opinions on nickel and LME, and where I think clearinghouse problems will come next.

Read more

This post will talk through the conceptual problem with profit maximising motives of a clearinghouse. Later posts will go through real world examples. The first concept to grasp is that the clearin…

Read more

Subscribe now

My observation on markets is that since 2016 onwards they have traded very differently to the past. Prior to 2016, I found that currencies, commodities, bonds and equities would trade in line with each other. Hence, if I had a strong view on any one of those asset classes, it would also inform my view on all the other asset classes. Furthermore, if the various asset classes starting trading disparately, it would be sign to be more cautious on my view. For many years, I wondered why this relationship broke down, with reasons ranging from central bank interference, to the invention of shale drilling and energy independence for the US. However, I now think the primary reason was regulatory reform that moved central clearinghouses to the centre of market pricing.

In financial terms, the dominant market pricing model when banks where the main clearinghouses was a Value at Risk (VAR) model. As risk started to increase, or was perceived to be rising, banks would cut their own risk exposure, and their exposure to other banks. This was forward looking, and also caused problems in one area to quickly cause risk to be pulled from all areas.

However, the new model, which uses initial margins is set by clearinghouses, who take no position in the market, but set pricing according to historic price action and liquidity. What this has meant in practice, is that risk is not repriced instantaneously through markets, but often with a lag. Tech offers a good example. Softbank is the largest VC investor in the world. It was also the premier Japanese dot com bubble stock back in 1999. As can be seen Softbank fell by nearly 50% in 2021, while Nasdaq continued to rise substantially. That is the risk that was being priced in Softbank was not being priced into Nasdaq. For an old trader like me, this makes me wonder if Softbank a buy, or is Nasdaq a short? This uncertainty made it difficult to hold any position.

My ongoing series on clearinghouses gives me the mechanics of understanding how this change in pricing system works. For those interested, I recommend becoming a paid subscriber just for the next month (£35) as I have at least another 4 posts on clearinghouses coming out describing in detail the problems with their risk pricing models. This will also encourage me to keep researching this area.

One very clear example of how risk pricing has changed from a forward looking model to a backward looking model is provided by Russia. Russia/Ukrainian issues first came to a head in 2014 coinciding with the Maidan Revolution. If we look at the CDS of Russia and Mexico, both commodity exporting nations with a chequered financial history, we can see that Russian CDS began to widen versus Mexico before the annexation of Crimea. That is markets were forward looking and already adjusting risk. However in 2022, despite the massing of troops on the border, Russian CDS failed to signal any risk, until the actual event happened.

What does that mean for markets today? Well in the US, before almost every bear market and recession, we have seen a spike in mortgage rates and a spike in oil prices.

Normally, such negative moves would lead to weakness in corporate credit markets - but so far high yield is at levels still associated with bull markets.

High yield reminds me of Russian CDS, being priced of the experience of the last ten years, but missing the forward looking view of rising energy prices and mortgage rates. The clearinghouse model, with its backward looking pricing allows us to observe changes before it gets priced in. Markets will likely work out this mispricing soon, so in my view now is the time to short.

Subscribe now

In my previous note, we highlighted the different experience of clearinghouses to two different bankruptcies, Lehman and Einer Aas. Lehman saw no losses at clearinghouses despite being of a much large scale, while Einer Aas required other members to recapitalise the clearinghouse.

The key is that under Lehman, the clearinghouse has plenty of initial margin on hard, while with Einer Aas, initial margin was plainly insufficient. Lehman when bankrupt under the pre-crisis model, while Einer Aas went bankrupt under the new model.

In the pre-crisis model, all banks are in essence acting as a clearinghouse, which requires them to require margin from all counterparties. Bank of America would have been a counterparty to Lehman at the time, and starting in 2007 the CDS of Bank of America began to rise.

Read more