The Chain

In the 90s, actuaries (and actuarial science) were at a bit of a low ebb as all kinds of public and private institutions had issues with pension liabilities. The actuaries – whose job it was to make detailed predictions of the likely long term demands on these funds and compare these to what the funds actually had in them – had taken a decades long jag of Panglossian assurance that all was for the best in the best possible of all worlds.

One of the odder things they did was to allow fund managers to claim that equities (stocks and shares) were more valuable than any other kind of investment (say government bonds, commodities, derivatives…). This was an accounting convention which allowed “profits” for returns on these equities (dividends or profits from sales, and gains in realisable value during the inflationary years) several years in advance – thus reducing the net liabilities of the pension funds.

By the mid 90s the flaws in this approach were beginning to show – and by the time of the equities crash in the early 00s (that’d be the dot-com bubble bursting) it was clear that something had to be done, and pension funds began de-risking – selling equities (at a loss) and using the proceeds to buy less risky products with smaller, but near-guaranteed returns.

Or so you’d think.

What actually happened in many cases was – having been burned by equities, and egged on by the interests of the defined-benefit pension-holders themselves – pension fund managers sought newer forms of investment that could offer a comparable return but at a lower risk. And a product was there to meet their needs.

David X Li, another actuary, began his financial career investigating ways of quantifying risk. Drawing on the copula function (specifically Gaussian Copulas) he came up with a way of predicting correlation between financial securities. Basically, he could turn the correlation between the likelihood of mortgage holder A defaulting and the likelihood of mortgage holder B defaulting into a single number – simple enough for investment bankers to understand.

Even though Li specifically warned against it, this (comparatively) simple output from a (largely poorly understood) formula became the driving force of the market for collateralised debt obligations – CDOs. (basically a CDO piles a whole bunch of debt into big box, and then slices it into segments rated at different levels of risk. You can buy a low rate of return with little risk of default (the “senior tranche”), or a high rate of return with a much larger rate of default. Li’s formula supported the generation of synthetic CDOs (sometimes CDO-squared), wherein the more-difficult-to-sell lower tranches were put into new boxes and resliced into tranches, thus generating new “senior trances” that could be sold.

And that -it has to be said – ended badly for the pensions of the world.

So where next for your savings and pensions? What about something that combined the the volatility and risk of equity with the anonymity, fungibility and algorithmic obfuscation of the CDO? What if we could combine these features with an entire lack of regulation or state backing, and – just for fun – a huge user base in illegal activity? Ready to fill your boots?

Such a product already exists.

There is significant global debate as to what species of financial instrument a “blockchain-derived currency” really is – a debate that has implications for taxation and licensing by governments.

  • Certainly it is tradeable against other currencies, with a widely fluctuating exchange rate – though it is backed neither by commodities nor a government/central bank.
  • With no extrinsic value, and no value backed by a known trustworthy party, it is clearly not itself a commodity, although it behaves as such.
  • It is issued by what could be considered a decentralised autonomous organisation in return for (computational) work conducted – work that is essential to the continuing viability of the organisation. But unlike equity it offers no voting rights or investor protections.
  • There is even some discussion as to whether or not it is a derivative.

What blockchain-derived currencies do offer is a lower administrative cost for transactions, where these transactions are simple. For on-network transactions there are no direct costs – though obviously the conversion to another currency at either end has a cost, and the “hidden” costs of power, connectivity and processing time are not factored in.

In terms of security, blockchains prioritise anonymity and encryption over direct trust. Indeed, “trust” is something of a dirty word – the industries that blockchain is slated to disrupt are generally those based on the need for trust.

Considering say, a bank: it offers some anonymity (it shouldn’t disclose the contents of my account to anyone without my permission), and some transparency (I am able to read about the strategy of the bank at my leisure, and monitor activity via publication and meeting). But the main reason to use a bank is for one of trust – I can (generally) assume that a bank will not do anything to jeapordise the money I have invested in it, and if this trust is misplaced I can (generally) assume that a central bank or nation state will ensure that my losses are mimimised.

By amping up the anonymity (it is impossible for anyone to know the content of my account, or link it to me) and transparency (all transactions are publicly listed), blockchain technology removes the need for trust – for all I (or indeed anyone) knows, Satoshi Nakamoto could be anyone or anything, and the other blockchain participants could be thieves and scoundrels or possibly even libertarians. The Ayn Rand forum accepts bitcoin payments – *shrug*

Bitcoin – to use the most common example, is a deflationary currency. There will only ever be 21m bitcoins – as transactions become more widespread the intention is to stop rewarding “mining” (the generation of the crypographic hash that is used to publicly record a transaction) directly with coins and instead allow miners to receive transaction fees. This change will happen automatically and was designed in to the bitcoin system

So – like CDOs – a smart algorithm has removed the need for me to trust anyone, and – like the old accounting convention around equities – an implication of guaranteed growth and return is undermined by deflationary reality. All of this is governed by the design of the algorithm – requiring (ironically) that we trust only in the wisdom of an anonymous coder that the various variables and correspondences have been set correctly.

Cue myriad screams of “this sounds AWESOME, how can we apply it to edtech?”

When silicon valley sneezes, edtech catches a cold about three-to-five years later. So if you’ve recovered from the “year of the MOOC” (yep, every year since 2012 it seems) then brace yourself for the first of many years of the blockchain! Hurrah!

Blockchain has provided a handily fashionable hook to attach increasingly moribund ideas like micro-credentialing and badges to. Of course we have to actually trust the issuers of the badges in questions, ensure that badges can be rescinded if this trust breaks down on either side, and despite this somehow have a means to incentive miners to keep growing the chain.

It’s fair to say that the blockchain is as poor a fit for micro-credentials as learning itself is – though this has not stopped various initiatives to combine the two. Despite this, there are genuinely interesting uses of blockchain technology which may have educational resonance: Ethereum is one that I see as being worth keeping an eye on more generally, and Mark Johnson is (as usual) thinking rich and stimulating thoughts.

Audrey Watters is researching this area in more detail, and I would highlight her ongoing struggles as a means of staying abreast of the tide of distributed anonymised ledger based nonsense that this year will bring.



15 thoughts on “The Chain”

  1. I think care needs to be taken with the distinctions. I don’t think anyone is excited about blockchain because it’s a CDO replacement. CDOs existed because banks declared them to exist. Banks can declare almost anything to exist! But they have nothing to with blockchain.

    With blockchain there is no declaring authority. Trust in it is upheld by its own socioeconomic dynamic: buying, selling, mining. In other words, it has AUTOMATED ITS MANAGEMENT. That’s what’s important. It should be seen in the same light as previous attempts to automate management – particularly Gordon Pask (management of learning) and Stafford Beer (automation of government).

    The real question behind blockchain in education is ‘Can we make educational management redundant?’ I don’t know the answer -but I’m not convinced it’s impossible -and given the awful and destructive behaviour at the top of institutions, its worth taking seriously.

    1. I think there’s more to be gained from seeing blockchain derived currency as a novel financial product. Those investing in bitcoin mining (and thus powering the entire ecosystem via their investment in specialised hardware) are seeing it in entirely this way. At the moment bitcoin provides a guaranteed return on investment with a low and manageable risk, just like CDOs used to.

      The second parallel is one of fungibility – miners process one or more transactions and it doesn’t matter where they come from. Every transaction is deemed to be equal, even though this isn’t the case. Currently (like in the early days of CDOs…) this assumption doesn’t matter – but bitcoin has a crash designed into it, where miners will (apparently – no one has tried this) make money from transaction percentages. Suddenly a small transaction or a cancelled transaction – which still must be processed – is worth a lot less.

      I’m not buying the automated management voodoo – someone (an anonymous someone) has written a set of instructions which is followed by a dumb machine. This is automated management in the sense that a robot in a factory is automated management… we’re misassigning the management role.
      The “manager” of bitcoin is the originator of the software, though he/she/they/it have declared that they have no wish to make further interventions (like the libertarian dream of a central bank). So in a sense, the currency is now unmanaged.

      Can we make educational management redundant – yes (by glorious and bloody revolution!!). But management by algorithm just obfuscates where the real decisions are taken (again, the CDO and equity investment parallel – or for an eduspecific example, look at Prof Grimm’s terrible experience…)

      1. In the end management is what does it does – if it kills the thing it’s supposed to be managing (which is a phenomenon which we know too well from human management!) – it’s a cancer.
        There are of course pathologies in BitCoin because there are greedy people, but for all the opportunities for greed, I don’t think anyone quite knows what they’re dealing with – they’ll take risks and they’re likely to get burnt. Will that kill it? I doubt it. BitCoin seems to have infinite lives. And automatic management.
        I think we are already seeing automatic management in some forms – albeit in rather nasty corporate wrapping (so there is an elite taking the profits!). Uber is the best example: a very simple algorithm which the drivers like – basically, who is nearest the next job. In expressing their appreciation of this, the drivers report something like:
        “when the taxi base allocated jobs, it wasn’t fair. Basically some people were in favour with the boss and got the jobs and others weren’t. With Uber the rule is simple – if I’m closest I get the job”
        When it comes to management, people can be much dumber than machines! Uber’s management merely acts to uphold the status of Uber, just as Facebook’s management upholds the status of Facebook. And the management of the university of …

        How about creating a peer-to-peer distributed-ledger oriented Uber where the profits are shared among the drivers rather than going to California? Seems feasible to me.
        And revolution? well…

Leave a Reply

Your email address will not be published. Required fields are marked *