By George Samman, the former CMO of Fuzo which is using blockchain to bring financial inclusion to the developing world, co-founder BTC.sx, a former Wall Street Senior Portfolio Manager
This post will explore the 2009 subprime mortgage crisis and the hypothetical impact a blockchain may have had with regards to the proliferation of toxic synthetic Mortgage Backed Securities (MBS) and Collateralized Debt Obligations (CDOs).
Hindsight is always 20/20 so it is not the intent of this post to say blockchain would have been the cure and stopped it from happening. Instead, it may have lessened the impact greatly by flattening out boom/bust cycles. When thinking of mortgage backed securities and asset backed securities this will be written from the lens of supply chains and provenance. This post will describe the assets involved, what provenance is, the subprime mortgage crisis and why distributed ledgers would have been instrumental in lessening the impact of the crisis. The global economy stood on the precipice of a global depression as the credit markets froze.
What Is Provenance?
Provenance refers to the tracking of supply chains. It’s important for purposes of integrity, transparency and counterfeiting that this asset and its path be known every step of the way. A supply chain refers to the creation of a network in which an asset is moved, touching different actors before arriving at a destination. The tracking of this asset has big value to the actors involved for the reasons mentioned above. This applies to financial synthetics such as asset backed securities (ABS) as much as anything else in the world considered to be an asset including: Norwegian salmon, diamonds, prescription drugs or Letters of Credit (LOCs) and Bills of Lading (BOLs). In fact due to the liquidity created out of these financial synthetics and the leverage built into them in as they move through the financial supply chain it becomes necessary to ensure two things: 1) Note how individual assets combine to form a newly made synthetic asset and 2) track possession of the asset as it moves through the supply chain. There are two kinds of medicine: preventative (measures taken to prevent the onset of a disease) and curative (treating a disease once you already have it) in this case a blockchain could be a preventative type of medicine.
In order for data provenance to be effective from a technology standpoint it must fulfill certain requirements & characteristics that would upgrade supply chain management and monitoring. According to Sabine Bauer in his paper titled “Data Provenance in the Internet of Things” they are:
1. Completeness: Every single action which has ever been performed is gathered.
2. Integrity: Data has not been manipulated or modified by an adversary.
3. Availability: The possibility to verify the collected information. In this context, availability is comparable to auditing.
4. Confidentiality: The access to the information is only reserved for authorized individuals.
5. Efficiency: Provenance mechanisms to collect relevant data should have a reasonable expenditure.
In addition to these requirements are some additional characteristics such as Privacy of individuals personal data. This flows into another characteristic which is Unlinkability, a state where this personal data must not be able to leave the system and get into the wrong hands. It needs a secure wall built around it. Finally it also have Linkability, which is simply total transparency of the chain and total traceability of the data particularly when it comes actions and modifications to it. A blockchain contains all of these requirements and as will be shown when the example of the 2009 subprime crisis is explained, it has additional characteristics which if applied to MBS and CDOs would have buffered the crisis in a way that no database could.
For provenance to work really well on a blockchain, the whole ecosystem (all parties involved) needs to be a part of the ledger so that network effects make for a much more robust supply chain and the associated integrity of the data that comes from it. It will also allow for easily identifying errors in the supply chain and a real time look at the data which is extremely important. Trade finance provides an example of the network effects and particularly when using Letters of Credit (LOCs). A Letter of Credit is defined as:
“A letter of credit is a letter from a bank guaranteeing that a buyer’s payment to a seller will be received on time and for the correct amount. In the event that the buyer is unable to make payment on the purchase, the bank will be required to cover the full or remaining amount of the purchase.”
While a blockchain would be valuable if you just had the buyer and seller using it to save costs and track payments and goods, it would remain incomplete and inaccurate at certain data points. (the same will be true for subprime mortgages and the making of them.) What does make this more powerful is to have the customs authorities, the banks, the big corporates, the shippers, and the manufacturers on the ledger. In other words, anyone involved in the chain of custody as the asset is moved either physically or digitally providing all parties with the ability to watch and update and confirm that the asset is genuine and moving along the chain to its intended destination. Another problem with LOCs which will also reared their ugly head in the MBS crisis, is the document heavy (mostly paper) credit processes that still remain intact. The corporate databases which are involved suffer from incompatibility and often high transaction costs to get them to communicate with each other. This makes the current system around supply chain management hard to maintain on a real-time basis. Physical items become access points to the ledger which can be identified, located and addressed by unique characteristics.
The 2009 Subprime Mortgage Crisis
A few definitions will be needed to explain the types of assets that caused the crisis. This will also be useful since provenance and distributed ledgers (as described above) would have been important in stopping the severity of the crisis.
Note: There are different categories of Asset Backed Securities (eg auto loans, student loans, leases, etc) The focus of this piece will be on Mortgage Backed Securities and Collateralized Debt Obligations. As theFinancial Crisis Inquiry Report noted (and found on Wikipedia) “The CDO became the engine that powered the mortgage supply chain.”
- Asset Backed Securities: a security whose income payments are from and collateralized (backed) by a specific pool of underlying assets. This pool usually consists (but doesn’t have to) of a group of illiquid assets which can’t be separately sold. These assets are then pooled together and a synthetic asset is made that can be sold to investors. This process is known as securitization and is supposed to “de-risk” the assets by diversifying them by bringing together tiny pieces of assets from a what’s supposed to be a diverse pool of underlying assets.
- Synthetic Asset: A mixture of assets that, when combined have the same effect and value as another asset. This also gives it the same capital gain potential as the underlying security.
- Mortgage Backed Security: a type of asset backed security that is secured by a mortgage or a pool of mortgages. The mortgages are sold to a group of individuals (a government agency or investment bank) that securitizes, or packages, the loans together into a security that investors can buy. The mortgages of an MBS may be residential or commercial. The structure of the MBS may be known as “pass-through”, where the interest and principal payments from the borrower or homebuyer pass through it to the MBS holder, or it may be more complex, made up of a pool of other MBSs. Other types of MBS include collateralized mortgage obligations (CMOs, often structured as real estate mortgage investment conduits) and collateralized debt obligations (CDOs).
4. Collateralized Debt Obligations (CDO’s): are securities backed by debt obligations that encompassed mortgages and MBS. Like other private label securities backed by assets, a CDO can be thought of as a promise to pay investors in a prescribed sequence, based on the cash flow the CDO collects from the pool of bonds or other assets it owns. The CDO is “sliced” into “tranches”, which “catch” the cash flow of interest and principal payments in sequence based on seniority. If some loans default and the cash collected by the CDO is insufficient to pay all of its investors, those in the lowest, most “junior” tranches suffer losses first. The last to lose payment from default are the safest, most senior tranches. Consequently, coupon payments (and interest rates) vary by tranche with the safest/most senior tranches receiving the lowest rates and the lowest tranches receiving the highest rates to compensate for higherdefault risk. As an example, a CDO might issue the following tranches in order of safeness: Senior AAA (sometimes known as “super senior”); Junior AAA; AA; A; BBB; Residual.
Banks try to make these tranches even safer by insuring them for a fee. This is called a Credit Default Swap.
These securities became really popular during the US housing bubble as housing prices soared and with it the amount of new mortgages (whether primary, secondary or tertiary). Any individual whether credit worthy or not was given a mortgage and this created a whole new pool of mortgages to package and repackage. The banks loved them because they could get these mortgages off their books for cash by passing them through. They were a way for investors to get a good rate of return in an extremely low interest rate world at a time when there incredible global demand for fixed income and these were seen as safe assets because the rating agencies were putting their seal of approval on these securities. However, even though the banks were getting these mortgages off their books the CDOs that synthetic CDOs that were made expanded the impact of what would come later: mortgage defaults. Before the creation of Credit Default Swaps (CDSs) and CDOs you could only have as much exposure to non-prime mortgage bonds as there were such mortgage bonds in existence. The extreme leverage used by the banks caused an “infinite” amount of synthetic CDOs to be created.
The banks, investors and everyone else involved in the financial supply chain had no transparency into their own risk management. There was (and still is) an inordinate amount of complexity in the securitization of these assets. As these products were really major money makers for banks, there was enormous pressure to beat your competitors in getting your products out quicker and faster, which led to sharp declines in underwriting standards. The securitization piece was happening off balance sheet and being guaranteed by the issuer. This caused enormous amounts of leverage to be used without an afterthought for how unstable the capital structures would become. Credit risk was extraordinarily underpriced. A recipe for disaster.
Another key domino was trust in the rating agencies. They were putting investment grade ratings on subprime tranches. Not only this, some of the AAA tranches contained nothing but subprime loans. The rating is very important for investment purposes as certain types of investors (pension funds, global municipalities, mutual funds, etc) have mandates on what they can purchase and these are generally only AAA rated. The CDO managers weren’t forced to disclose what the securities contained, because the contents of the CDOs was subject to change. The underwriters who structured them were less concerned about the ingredients, the rating is what mattered most.
In other words, the default rate became 100% as all the MBS and CDOs began blowing up. Real people with real homes who were actually able to pay their mortgages watched the value of their houses drop dramatically versus what they paid for it and people just started walking away. To add to the complexity, in many cases, not even whole mortgages were out into these synthetics but tiny slices of mortgages were put into different synthetics so chain of title was completely lost in this process. To add to the complexity these mortgages and pieces of mortgages were geographically dispersed. In order to de-risk they were packaged together and sold.
Aside from non-credit worthy individuals being able to own homes as less and less stringent standards were used robosignings were another byproduct and apparently are still being used to this day. Robosignings refer to a practice where a signature from a bank or mortgage official on legal documents guarantees that the information is accurate. These were happening without verifying the information, or signatures being forged and overall not complying with notary standards.
This system that is in place to deal with this chain of custody is Mortgage Electronic Registry System (MERS). It is a private company which owns an electronic registry that’s purpose is to track servicing rights and ownership of mortgages in the USA. An obvious prerequisite is that the information must be digital. (which was not the case back then by a long shot and is still not the case today.) They are supposed to track chain of title but failed at this and instead foreclosed on millions of homes in which they did not even have proper paperwork, making decisions they should not have made (since they did not own the loans have have transparency into them) They were supposed to hold title nominally so that buying and selling of MBSs could be made easily without registration of ownership changes with local governments. (Yes you read that right!)
Synthetic assets created on top of synthetic assets insured by credit default swaps and a real asset underlying. Only a handful of people understood what these were and the amount of leverage they created. There was no way to control risks or even truly understand what they were. The definition of ownership came into question in many ways. When everything started to get unwound mortgages were lost, who owned the mortgage was questioned, who was responsible for paying was asked. Fake signatures and false registration documents were rampant. LLC’s signed for sub-LLC’s which signed for sub-sub LLC’s. This same sub system applied to mortgage companies as well and most of the paperwork here was lost stolen or destroyed. This was well documented. There was no way to trust a lot of the data and it was being found that the same collateral was being passed around for different loans. (rehypothecation) The middlemen had all gone bust so access to the money was lost and where it was even located. The end lenders had major problems. When all the asset backed securities blew there were still real mortgages behind them that needed to be dealt with (or not) but the sheer complexity involved made this a most difficult task.
Note: In the case of financial asset backed securities: Distributed Ledgers should be used.
- Blockchains are a way to introduce transparency into supply chains and create entirely new opportunities for participation. This shared, secure record between parties would break supply chain data out of its silos. Imagine a ledger which included all the parties involved in the subprime mortgage crisis being a part of: Investors, underwriters, asset managers, trustee and collateral administrators, accountants, attorneys and regulators.
- All parties would be able to see what made up each and every tranche and each every MBS and CDO. Who bought it and sold and whom to along the supply chain. Each party could get a real time view into their risk profile and adjust it accordingly. This includes counterparty risk, credit risk, default risk and ownership risk. By having all participants on the ledger a real time rating system is established as opposed to the way it is done today in which it lags behind and is semi-static. By having all the participants on the ledger and allowing real-time read/write functionality, risk will fall significantly and compliance standards will rise dramatically (tragedy of the commons).
- Smart contracts built on the ledgers will self validate the authenticity of identity, the mortgage agreement itself, the chain of custody of the titles and automate payments of interest and principal between parties. By having a real-time view of payments, imagine the possibilities. These records and registrations are created in a distributed way (by having all parties on the ledger) which creates new ways to track identity and the reputation of all involved. This creates real time ratings for the assets involved and the counterparties on the ledger.
- Blockchains can create a formal registry to identify each to identify each individual mortgage or slice of a mortgage (or any asset) and track chain of title through different exchange points in the supply chain. (Goodbye MERS)
- When blockchains track the movements of assets from beginning to end and all points of exchange in between they can also see how these assets become repackaged and continue real time tracking. This allows all users to know what exactly went into the creation of the asset and how it has changed along the way.
How powerful is this for custodianship and prosecuting bad actors. Better yet, stopping them from acting bad in the first place. Imagine if regulators could have tracked back all the MBSs and CDOs as well as who owned the mortgages and who the counterparties were. This would have made enforcement actions much easier to attain. It would have made reckless behavior from all parties much more difficult and it would have precluded many investors from buying these assets to begin with. There would have been transparency into asset structuring and asset pricing.
In the aftermath, companies were forced to merge (Wachovia, Washington Mutual, Countrywide, US Bank etc) with the bigger banks in super quick fashion without having any view into the risks they were inheriting. They did not even know where the data was stored and in what form. Forensic experts were brought in to try and peel through the layers of the onion and uncover many different information points, ascertain ownership, find out who the counterparties were and unwind businesses. It was extremely disorderly and is still beginning done.
If anything this should be a statement to the fragility of the financial system and the technologies in place at present. Complex systems like the financial system need robust solutions. Blockchain is a robust technological system which has many answers for dealing with this the use case presented. In the future financial calamities may not be able to be stopped but they can be tempered.
The article first appeared in George’s Blog