This is a funding / liquidity issue and not a risk or solvency issue.
LDI says I have 30 year rate exposure, if rates tank the pv of those liabilities explodes, it drops if rates go down, therefore hedge by selling floating for fixed. The problem is that even though the decrease in their liabilities from rate increases is definitely a 'real' gain for their balance sheet, it's not cash, so they can't post anything against their mark to market losses on the financial assets / derivatives. This is the same as what happened in commodity markets a few months ago where people owned physical against short derivates and couldn't post margin. In a properly functioning market, someone should happily lend the pension the cash they need for collateral, because the balance sheet is fine. Because liquidity is extremely tight and markets are generally broken, pensions are perfectly solvent but completely illiquid and have to firesale assets until they actually are insolvent.
US insurance cos are similar in 'real balance sheet' terms, but get to hold all their bonds at book value and try to structure hedges to not be marked to market or they might not hedge.
When you're in London, please go see the UK government and explain this in terms as simple as you use in this post. It would be news to them. Also helpful. Great work as always Alf.
as a LDI related hedging strategist, I think the whole thing needs a regulation overhaul, so that current financial plumbing will match with the current high inflation high interest rate environment. Also, for pension fund, their liabilities are more sensitive to the interest rate changes than their asset, so they have to use somewhat liquid instruments to dynamically adjust their hedges, and a standardized interest rate swap fits that bill.
One thing you mentioned, but I haven't seen mentioned enough with respect to the UK vs. US, is that the UK regulator virtually mandated LDI for DB and DC plans, and UK plans often have inflation linked benefits, so the UK pensions dominated the linkers (ILB) market. In the US LDI isn't nearly as prevalent as it is in the UK (for various reasons) and not as many US plans have inflation linked benefits, so the hedging is more in the nominal bonds market vs. TIPS. So while it *could* happen in the US, it is a lot, lot less likely.
If it’s not fully collateralized with risk free assets it’s leveraged and reduces the size of move required to get offside. Throw in crowding caused by a huge cash cow for various enablers and a less liquid market. Not first time these guys have been smoked by derivatives…look at LOBOs in mid 2000s…
The dark horse in non LDI books will be illiquids and how the leverage used stands up in a tightening regime with an economy in a downturn. Marks are way off listed multiples currently…
Less likely in the US than UK for a few other reasons that you didn't mention:
1) US liabilities are generally nominal (no inflation linkage) and tied to long A or better credit for both accounting and funding rules. For a variety of reasons, US long credit market is of reasonably decent size allowing majority of LDI to be done in cash markets as opposed to derivative markets. Less of a "barbell" in strategy than typically seen in UK.
2) There are a few major players using swaps, but they represent a smaller fraction of LDI hedging and the highly-levered pooled funds never really took off to same degree as UK.
3) Despite being lobbied by the investment banks to "bless" swap overlays, the DoL which overseas pension regulation under ERISA, has been a little more circumspect. They've opined that better Asset-Liability matching is a good thing, but been reluctant about stating that using swaps to achieve a better match was a sage harbor strategy. (This was as of about 5 years ago.)
And right at the heart of the issue, like you mention, we're terrible at framing risk beyond parameters we've seen before....not sure if that's a solvable problem? Do we need to start incorporating 10+ sigma moves into VaR and other risk analyses..? Thanks, Alf - wonderful as usual.
Holding alot of USD, however will people be able to use their "deposits"/digital bank checks legally defined as equal to cash, if we see a deep and/or sharp enough turmoil in markets?
Could it be that authorities freeze peoples accounts making cash practically useless? I.e., you won't be able to buy anything when there is actual blood on the streets.
Money today is mostly bank deposits, which means your money is a liability of a commercial bank which is highly regulated. By definition you always run that risk.
1. true that as rates climb the fixed receiver swaps / long duration bonds go underwater but in a DB plan same goes for liabilities. so it is more of a liquidity issue than a solvency one. 2. when rates were in negative territory pension funds had the yield issue because rates were low! so they never get it right. 3. being basically an exercise of matching asset and liability duration why massive deviations are tolerated by boards, contributors etc...?
This was an eye-openings article. I was always curious on how pension funds truly operated.
Perfect timing as I was reading about Oxford Life Insurance business. Looks like their investments include 11.7% allocation in Mortgage issuance and a percentage in Bonds as well as some derivatives Call options on the S&P500.
These guys have no debt and use their full equity and returns from customer premiums.
My concern is their rating. They are rated “A” in the Best Credit Rating report. How are they rated “A” in this current economy? Especially with those call options on the SPX. It’s quite impressive.
When the swaps are bilateral, mostly the pension funds (at least in the NL) will be able to provide VM in money or in bonds. And bonds they have in abundance. CCPs only accept money as VM. As pension funds do not hold much liquidity, having to post money as VM is an issue. Exactly for that reason, pension funds until now have been exempted from the clearing obligation. Thanks to that, Netherlands pension funds hold most IRS bilaterally and a scenario like in the UK is less probable.
Since the GFC regulators have effectively prohibited banks from warehousing uncollaterized OTC exposures. Banks are heavily regulated to manage liquidity risk better than most however these exact risks have been pushed away from banks balance sheets due to mandatory clearing/margining.
I think should be inferesting an article comparing and analising equities AMD bonds in the 3 posible scenarios: hard landing, soft landing and no landing (stagflation)
I have some long bonds in the ETF portfolio but it should be seen in a portfolio construction framework where I am very short equities and risk assets too. As an outright long exposure I would wait a bit longer.
I'm still buying TLT, for whatever it's worth. Small pieces at a time, each time it drops further. I expect the catalyst for bond yields to start dropping is an acknowledged recession and fed pivot. Also, once inflation prints start coming in lower. I think the key here is if you want to buy TLT, you need to view it as a longer term play. Might start working in ten days, might take six months.
This is a funding / liquidity issue and not a risk or solvency issue.
LDI says I have 30 year rate exposure, if rates tank the pv of those liabilities explodes, it drops if rates go down, therefore hedge by selling floating for fixed. The problem is that even though the decrease in their liabilities from rate increases is definitely a 'real' gain for their balance sheet, it's not cash, so they can't post anything against their mark to market losses on the financial assets / derivatives. This is the same as what happened in commodity markets a few months ago where people owned physical against short derivates and couldn't post margin. In a properly functioning market, someone should happily lend the pension the cash they need for collateral, because the balance sheet is fine. Because liquidity is extremely tight and markets are generally broken, pensions are perfectly solvent but completely illiquid and have to firesale assets until they actually are insolvent.
US insurance cos are similar in 'real balance sheet' terms, but get to hold all their bonds at book value and try to structure hedges to not be marked to market or they might not hedge.
I agree! And I hope this was clear from the article as well :)
nice comment
When you're in London, please go see the UK government and explain this in terms as simple as you use in this post. It would be news to them. Also helpful. Great work as always Alf.
Thank you! Hope things are going well on your end :)
We are, rather surprisingly given the market and folks' prediliction to spend money on investment research during such times, on fire. In a good way!
as a LDI related hedging strategist, I think the whole thing needs a regulation overhaul, so that current financial plumbing will match with the current high inflation high interest rate environment. Also, for pension fund, their liabilities are more sensitive to the interest rate changes than their asset, so they have to use somewhat liquid instruments to dynamically adjust their hedges, and a standardized interest rate swap fits that bill.
Yep!
One thing you mentioned, but I haven't seen mentioned enough with respect to the UK vs. US, is that the UK regulator virtually mandated LDI for DB and DC plans, and UK plans often have inflation linked benefits, so the UK pensions dominated the linkers (ILB) market. In the US LDI isn't nearly as prevalent as it is in the UK (for various reasons) and not as many US plans have inflation linked benefits, so the hedging is more in the nominal bonds market vs. TIPS. So while it *could* happen in the US, it is a lot, lot less likely.
Indeed, that’s a good point
If it’s not fully collateralized with risk free assets it’s leveraged and reduces the size of move required to get offside. Throw in crowding caused by a huge cash cow for various enablers and a less liquid market. Not first time these guys have been smoked by derivatives…look at LOBOs in mid 2000s…
The dark horse in non LDI books will be illiquids and how the leverage used stands up in a tightening regime with an economy in a downturn. Marks are way off listed multiples currently…
Less likely in the US than UK for a few other reasons that you didn't mention:
1) US liabilities are generally nominal (no inflation linkage) and tied to long A or better credit for both accounting and funding rules. For a variety of reasons, US long credit market is of reasonably decent size allowing majority of LDI to be done in cash markets as opposed to derivative markets. Less of a "barbell" in strategy than typically seen in UK.
2) There are a few major players using swaps, but they represent a smaller fraction of LDI hedging and the highly-levered pooled funds never really took off to same degree as UK.
3) Despite being lobbied by the investment banks to "bless" swap overlays, the DoL which overseas pension regulation under ERISA, has been a little more circumspect. They've opined that better Asset-Liability matching is a good thing, but been reluctant about stating that using swaps to achieve a better match was a sage harbor strategy. (This was as of about 5 years ago.)
All correct!
And right at the heart of the issue, like you mention, we're terrible at framing risk beyond parameters we've seen before....not sure if that's a solvable problem? Do we need to start incorporating 10+ sigma moves into VaR and other risk analyses..? Thanks, Alf - wonderful as usual.
Convex moves have always been beyond the scope of many risk models…which is quite an issue
Holding alot of USD, however will people be able to use their "deposits"/digital bank checks legally defined as equal to cash, if we see a deep and/or sharp enough turmoil in markets?
Could it be that authorities freeze peoples accounts making cash practically useless? I.e., you won't be able to buy anything when there is actual blood on the streets.
Money today is mostly bank deposits, which means your money is a liability of a commercial bank which is highly regulated. By definition you always run that risk.
1. true that as rates climb the fixed receiver swaps / long duration bonds go underwater but in a DB plan same goes for liabilities. so it is more of a liquidity issue than a solvency one. 2. when rates were in negative territory pension funds had the yield issue because rates were low! so they never get it right. 3. being basically an exercise of matching asset and liability duration why massive deviations are tolerated by boards, contributors etc...?
All correct points!
On 3: because the alternative was to announce pension cuts :)
This was an eye-openings article. I was always curious on how pension funds truly operated.
Perfect timing as I was reading about Oxford Life Insurance business. Looks like their investments include 11.7% allocation in Mortgage issuance and a percentage in Bonds as well as some derivatives Call options on the S&P500.
These guys have no debt and use their full equity and returns from customer premiums.
My concern is their rating. They are rated “A” in the Best Credit Rating report. How are they rated “A” in this current economy? Especially with those call options on the SPX. It’s quite impressive.
I’ve included their Best Rating Report here: https://oxfordlife.com/about-us/pdf/AM_BestA.pdf
Again thanks so much for this valuable knowledge. Keep ‘em coming! Been learning a lot!
Glad you liked it!
When the swaps are bilateral, mostly the pension funds (at least in the NL) will be able to provide VM in money or in bonds. And bonds they have in abundance. CCPs only accept money as VM. As pension funds do not hold much liquidity, having to post money as VM is an issue. Exactly for that reason, pension funds until now have been exempted from the clearing obligation. Thanks to that, Netherlands pension funds hold most IRS bilaterally and a scenario like in the UK is less probable.
Hi Frans,
Partially correct. Many Dutch PF moved swaps to central clearing.
Since the GFC regulators have effectively prohibited banks from warehousing uncollaterized OTC exposures. Banks are heavily regulated to manage liquidity risk better than most however these exact risks have been pushed away from banks balance sheets due to mandatory clearing/margining.
Correct!
I think should be inferesting an article comparing and analising equities AMD bonds in the 3 posible scenarios: hard landing, soft landing and no landing (stagflation)
Great idea!
Do we continue to stick with the TLT trade? When should we accumulate more? And when should we expect bond yields to start dropping?
I have some long bonds in the ETF portfolio but it should be seen in a portfolio construction framework where I am very short equities and risk assets too. As an outright long exposure I would wait a bit longer.
I'm still buying TLT, for whatever it's worth. Small pieces at a time, each time it drops further. I expect the catalyst for bond yields to start dropping is an acknowledged recession and fed pivot. Also, once inflation prints start coming in lower. I think the key here is if you want to buy TLT, you need to view it as a longer term play. Might start working in ten days, might take six months.
Thank you Alf. I appreciate that you share your expertise with us. Sincerely - Thank you.
Welcome!
Great article - appreciate your work!
Thank you!