Capital, like water, tends to flow around obstacles. Try to dam its movement at one point, and slowly but remorselessly it will find its way around.”
This quote came from an article in The Economist a few years ago discussing the effects of increased standards, incentives, and regulations designed to help prevent the type of collapse asset securitization markets experienced in 2007 and 2008. It reminds me of one of my favorite books, What Technology Wants by Kevin Kelly which wonders whether the growth and spread of technology are almost inevitable and autonomous and if so begs the question what does Technology want next?
Asset Securitization is the process of taking a group of assets and through financial engineering transforming them into a security. A great read about the good, the bad, and the ugly facets of securitization is Can Securitization Work? Economic, Structural and Policy Considerations* by Timothy J. Riddiough. Trillions of dollars participate in securitization in both mature and emerging markets. Regulation for these markets has been a hot topic anywhere you look.
Regulatory frameworks are a time-tested tool to create obstacles to redirect capital away from portfolio risk concentration, over leverage, black box financial structures, and conflicts of interest that crop up in every volume business. But it is harder to find discussions on the role of technology within Securitization. We are in the middle of a technology revolution at a time when technology is capable of far more than automating a workflow to gain 15% efficiency.
Technology is fundamentally changing the way we work and live with new means of communication and collaboration and disruptive innovations in nearly every single industry. The good that comes from Securitization is well documented. Is it possible that some of the downsides are susceptible to technology disruption as an alternative or complement to increased regulation?
What Does Securitization Want from Technology?
1. Each Deal to Become Its Own Universe of Financial Data
Deborah Estrin asks What happens when each patient becomes their own “universe” of unique medical data? Securitization wants in on this action. The amount of data from disparate data sources that goes into securitizing a pool of assets, from loan production through new issuance modeling and analysis, the ratings game, and ultimately surveillance and monitoring post origination over the life of the deal is staggering. Consider one of the most popular data feeds LIBOR. Every lender onboards and maintains their LIBOR feed. Every Underwriting firm. Every Due Diligence firm. Every Ratings Agency. Every Counterparty. Every Investment firm.
For a single deal, 8 companies are maintaining their own copy of this single data feed. And for the deal there are probably 9 other data feeds with relevant information. We are talking about work equivalent to onboarding and maintaining 80 data feeds. All by companies whose expertise lies outside of technology. Harmonizing this data is a herculean task for everyone except maybe the large banks who have the largest IT budgets and in-place infrastructures.
This is just for one deal. Scale is achieved somewhat by the fact that each company participates in many deals. Traditionally technologists like to abstract and generalize concepts from the data and try to make every deal “look and feel the same” from a data perspective. It makes it easier to render charts. But every deal is not the same. Large deals from mature markets are totally different than small deals in emerging markets. Securitization wants each deal to come with its own private data lake of sorts that holds every single piece of data that the deal has ever touched with no abstractions present that could possibly dilute the data or the context. Securitization wants deals to hoard their own data.
2. Loan Production Efficiencies
Securitization wants ruthlessly efficient loan production channels. Securitization differs from traditional lending in that it seeks to gain efficiencies by handing off facets of loan production and servicing to specialists who theoretically are better at what they do than a large organization trying to handle every facet of production in-house. While it is true that specialists are likely more efficient in their role this model yields a very disintegrated production channel. With disintegration comes complexity since so many entities are involved and data\communication has to flow from one entity to the next in coordinated fashion. Securitization doesn’t care.
This is a people problem because we hate to share and there is usually no incentive to. Securitization wants technology to create highly integrated yet decentralized production and servicing pipelines that are strengthened by competition over an even playing field of access to the universe of deal data.
3. Transparency On Demand
Securitization wants transparency, however not some ideal notion of full transparency. After all, Securitization is a game that can be thought of to reward those who succeed in doing the work of uncovering the most vital and relevant information. But a certain amount of transparency is required to alleviate the panic that can be caused when market shocks affect a deal. The unavailability of data can make it difficult to evaluate the effects on a portfolio.
For more complicated deal structures it is difficult to find the real economic value within the underlying assets without access to the right data. Securitization seeks to alleviate these panic attacks by storing the underlying universe of deal data and being able to produce it to the right parties at the right times. There should be no technological barriers which prevent anybody from participating in surveillance. Traditionally only the biggest players have deep surveillance infrastructure, however, every participating entity no matter how small has a stake in surveillance should be able to access relevant data. The underwriters want to surveil too.
4. Surveillance Based On Active Experimentation
Securitization wants more people asking questions about the data. With the vast disparity in asset classes and loan product types no two deals should be surveilled in exactly the same manner. There are many facets to portfolio risk management:
- Survey: Making a single observation to measure and record something.
- Surveillance: Making repeated standardized surveys in order that change can be detected.
- Monitoring: Surveillance undertaken to ensure that formulated standards are being maintained (JNCC 1998).
- Recording: Making a permanent and accessible record of significant activities.
Securitization wants technology to move past the notion of passive reports (what happened yesterday) and towards the notion of active experimentation (what could happen tomorrow). Analysts do not know what questions they are going to want to ask about their deal tomorrow. They do not know what Survey should be repeated for all time to yield the most relevant watchlist. They need to be able to experiment; to form a hypothesis about the deal, test that hypothesis over a period of time and use the results to make fact-based decisions on the health of the deal and the processes by which they want to continue to surveil and monitor. The transition away from passive reporting to active experimentation requires technology that puts data in the hands of financial analysts, reducing or even eliminating dependence on IT departments.
If you participate in Securitization markets do you have access to all the data you need? Are there technological barriers that prevent you from surveilling the market? Do you rely on active experimentation or passive reporting? Do you want the same thing Securitization wants?
