A presentation to the Global Custody Forum, London 1 December 2015
A presentation to the Global Custody Forum, London 1 December 2015
More than a quarter of a century has passed since I found myself, one bleak autumn evening, sitting in the Lower Thames Street offices of the Trustee Savings Bank, interviewing the head of securities services.
It is so long ago that in those days people were still allowed to smoke in the office.
And the head of securities services at the TSB certainly did.
I had only one question for him, really, and it was this: was the TSB worried that the American global custodian banks were competing for his clients?
"There is," he croaked, drawing deeply on a State Express 555, "a lot of paranoia about."
Another seven years passed before Andy Grove, the CEO of Intel, published that now legendary work of business literature, Only the Paranoid Survive.
So the head of securities services at the TSB never got to read it.
But you can, and I urge you to do so.
Because I detect, in my conversations with the leaders of this industry, all of the corporate pathologies which Andy Grove identified in that book.
A refusal to take technological change seriously.
Have you had a conversation in the last few weeks or days with somebody in your organisation or in the industry along the lines of, `Oh yeah, and blockchain is going to end world poverty as well.’
I know that I have.
Do the older people in your organisation still believe that your huge technology budget is a barrier to entry to your business?
I know custodians who do.
Even in a world in which one entrepreneur tells me it will cost his investors just £5½ million a year – that is about one 180th of the annual technology budget of, say, State Street - to build and operate a settlement utility capable of settling 1 billion transactions a second.
Ours is an industry in which technology is no longer raising the barriers to entry: it is razing them to the ground.
Then there is the voice of experience.
The sound of those who have seen-it-all-before.
The sceptics who hold that scepticism is the highest form of business insight.
The generation, scarred by the Dot Com boom, who are quite convinced that blockchain is all sizzle and no sausage.
They sound wise.
They are not hard to find.
And they may even be right.
But what if they are not?
Deciding not to decide is still, as they say, a decision.
Perhaps you have observed some of your colleagues leaving banking for the FinTech industry, and not just because they cannot get a job in banking any more.
Perhaps you work for one of those banks visited recently by a blockchain start-up armed with a clever plan to revolutionise custody and settlement.
And the only response of your colleagues was that “turkeys do not vote for Christmas” or “he won’t want to hear that.”
Perhaps you work for that well-known global custodian bank that a year ago had identified a brilliant solution to its existential strategic dilemma.
And then decided, earlier this year, that it was not going to get into the CSD business after all – or not yet, anyway.
In business, as in life, it is hard to imagine a future that is different from the past.
It is hard to believe that the service we supply can be built or delivered in a different way.
And it is quite impossible to believe that neither we nor the company we work for will not continue in business indefinitely.
Yet it happens.
I could think without difficulty of 23 banks which have exited the global custody industry since I first started writing about it in 1989.
Note they include the TSB, highlighted in red.
Its clients were sold first to Lloyds Bank – then given away to State Street, in 1999.
Now none of those businesses actually died.
They got sold, bought, or merged.
What is different about what is happening now is that our industry is at what Andy Grove called a “strategic inflection point.”
It looks like this.
At a strategic inflection point, companies do actually die.
Because someone or something has created an entirely new way to deliver the service that we supply.
A “strategic inflection point,” in other words, is a fundamental shift in the underlying structure and economics of a business.
Now, the business custodians are in today is called “banking.”
It might not always feel like that.
After all, custody is often described as a fee-earning, off-balance sheet business.
But there are in fact a number of activities custodians undertake which are, on any reasonable definition of the term, “banking.”
Banking is, at bottom, accepting money belonging to one group of people and lending it out at a profit to other people.
If we look at the aggregate balance sheets of a trio of leading global custodian banks that is a pretty accurate description of what they do.
Three out of four dollars on the liability side consist of client deposits.
These are invested, on the asset side, in other deposits, money market instruments, securities and loans.
Since the client cash deposits are not invested dollar-for-dollar in cash deposits, it is obvious that global custodian banks are also engaged in another activity characteristic of banking: maturity transformation.
But balance sheets do not tell the whole story.
Custodians lend money to clients to facilitate settlement.
They lend securities, so other clients can secure credit.
They execute equity and FX trades for clients.
Often as principal.
In fact, if we add up the value of the banking activities of the two largest global custodian banks, they account for a third of the revenues.
But what about the other two thirds?
It is made up of the various fees that custodians ostensibly collect without putting the balance sheet at risk.
I say “ostensibly” because it is pretty obvious that these activities do put the balance sheet at risk.
Custodians may have given up indemnifying commercial paper-funded conduits, which created a string of embarrassing losses in the crisis.
But they still indemnify securities lending clients against the loss of their assets.
Nowadays, they also indemnify the same clients against losses on the reinvestment of cash collateral in the repo market.
Now, those risks are collateralised, but they are not on the balance sheet.
And they are large numbers.
BNY Mellon currently records a total of $365.5 billion in securities lending indemnities: a sum equivalent to 97 per cent of its total assets.
State Street records $330.8 billion, a sum equivalent to 130 per cent of its assets.
Custodians themselves enter swap agreements where they are only a credit rating adjustment away from having to post hundreds of billions of dollars of additional collateral.
And even in their core custody business – the citadel of the off-balance sheet, fee-earning franchise of the industry – custodians are exposed to a string of open-ended risks.
Every securities and foreign exchange settlement represents a risk of non-payment.
Every corporate action missed is a potential claim on the balance sheet of the bank.
Every proxy vote missed or miscast is a law suit in waiting for consequential damages – perhaps even direct losses.
Thanks to FATCA, and global FATCA, and Section 302, even a tax payment withheld or claimed in any one of a hundred different jurisdictions is now a potential source of an official penalty or a law suit from a client.
Two years ago, banks paid $1 billion in fines for getting tax wrong.
So far this year, they have paid eight times as much.
Above all, every sub-custodian represents a combination of counterparty, operational, market and jurisdictional risks.
Every CSD and every CCP contains within it the risk of the mutualization of losses if another member fails to honour its obligations.
There was a reason global custodians preferred not to indemnify their clients against losses at the market level.
But now they do - indeed must.
I could go on.
But I think I have said enough to prove that, if banking is at bottom the management of risk, global custodian banks are heavily invested, in everything they do, in the business of banking.
The risk they take is that of losing the assets that belong to their clients, and having to make them whole.
Since so many of those risks are not recorded on the balance sheet, it is tempting to describe them as offering an unreserved insurance policy.
Certainly, an insurance policy is what investors think they are buying.
Crudely speaking, they do not care what they invest in, or where, because they expect their custodian to make them whole.
But investors have misunderstood the nature of the trade they have made.
What custodians really do is re-distribute risk from the banking markets to the collateral markets.
In this sense, they are the shadow bankers par excellence, making the cash and securities of risk-averse buy-and-hold investors available to risk-hungry investment banks and money managers.
Now the regulators have made clear that they do not regard shadow banking as a socially useful activity.
They have forced investment banks to exit the hedge fund industry.
They have increased regulatory capital requirements.
Suppressed maturity transformation.
Re-regulated money market funds.
And forced banks to hold a higher proportion of high quality assets.
They have set about gathering data on the securities lending and financing markets, and the OTC derivative markets too.
Measures of this kind are causing direct and sustained damage to the revenues of the global custodian banks.
Securities lending: down 70 per cent.
Foreign exchange: down 30 per cent.
Net interest revenue: static.
On deposits up one and a half times.
It is now clear that the downturn in earnings from securities lending, FX and cash is not cyclical.
It is part of a structural change in the environment in which this industry operates.
In effect, the regulators have re-written the ratio between risk and reward in the global custody industry.
And this is what it means.
This chart shows the return on equity achieved by each of the six main business segments of J.P. Morgan between 2004 and today.
From the first quarter of 2005 until the third quarter of 2009, the custody business outperformed every other business division of the bank.
Only the other off-balance sheet activity of the bank – asset management – could get close to custody in terms of return on equity.
Once custody was merged with the investment bank in the second half of 2012, returns shrank dramatically.
Now, it could be that between 2005 and 2009 the leaders of the custody business at J.P. Morgan were doing a fantastic job.
Much as I admire them, I doubt that explains this chart.
What they were really doing was systematically under-estimating the risks of being in the custody business.
That was not a mistake.
The mis-pricing of risk was deliberate policy.
The purpose of being in custody at all was to generate revenue without allocating capital.
The idea was not to make money from custody, but to compete on price to get assets into custody, where they could be exploited or re-used - in FX, in cash, and in stock loan.
All major global custodians pursued the same policy.
It is why, despite acquisitions, core custody fees have flat-lined for years.
Wishing away the risks assumed in custody agreements was always unwise.
Now it is dangerous.
For the regulators have raised significantly the cost of mis-priced risk.
AIFMD has extended liability for losses to assets held by prime brokers.
UCITS V will extend it to CSDs.
On some readings of European law, custodians are now liable to make clients whole even for indirect losses, such as loss of profit if a transaction fails to settle on time.
Lawyers will doubtless get well paid to debate these matters.
But the practical reality is that in Europe custodians should now regard themselves as strictly liable for any and every conceivable loss their clients might incur.
And what is happening in Europe today will become the norm elsewhere as other clients demand the same terms as AIFs and UCITS funds are able to command.
If custodians mispriced risk before, they are certainly mispricing it now.
For they are insuring everybody against everything.
The historical trade-off –assets in custody in return for the right to re-use those assets – no longer works.
That is one reason why this industry is now at a strategic inflection point.
But there is another reason.
This is technology.
By technology I do not mean that blockchain is going to change the economics of the securities services industry overnight.
Clearly, it is going to have some impact.
Because it has the defining characteristics of every disruptive technology.
That is to say, it currently performs less well than the incumbent alternative – particularly in terms of speed.
It is also used only by a small class of enthusiasts on the fringes of the industry.
Above all, blockchain promises a simpler, less expensive and more convenient way of delivering mainstream custodial services such as settlement and asset servicing.
But the truth is that blockchain is just one expression of a deep underlying principle of nature.
This is that anything physically possible is in principle computable: all that needs to be applied to render it in reality is a sufficiency of memory and processing power.
And processing power.
And memory capacity.
They are all growing, not at linear rates, but at exponential ones.
According to Ray Kurzweil, we will by 2050 be able to buy for just $1,000 an amount of computing power equivalent to that of all the brains of all mankind.
If so, you won’t be offshoring your back office to Bangalore any more.
You will be giving the work to robots.
In other words, the disruptive technology this industry faces is not blockchain.
It is artificial intelligence.
And while I do not doubt that this industry attracts some of the finest minds of our civilisation, we will not need particularly clever robots to do global custody.
With or without blockchain, everything we do in this industry can already be done by digital technology.
Not just better.
Not just more cheaply.
But effectively for free.
So the disruption of this industry by technology is going to happen.
The only question is when, and how.
To answer it, let me hazard some predictions about what zero marginal cost computing might to do to this industry.
And offer some suggestions as to how custodians can main relevant in a world of robo-custody.
In the past, investors valued the intermediation of a bank balance sheet between them and the borrower – and were prepared to pay a spread for that reduction in risk.
In future, they will not.
Instead, digital platforms will solve the problem of risk by assembling tens of thousands of borrowers.
Each of them credit-scored, using hard data based on audited accounts, tax returns, locations, lifestyle and spending patterns.
Instead of bank treasury departments recycling deposits into deposits, algorithms will diversify risk across tens of thousands of borrowers.
So deposit-taking goes.
But credit assessments do not.
Custodian bank treasury departments can reinvent themselves as credit monitors - aggregating and analysing the data necessary to assess counterparty risks properly.
Money market funds will go the same way.
Because money managers, like bank treasurers, will in future add nothing but cost.
Secondly, foreign exchange.
For decades, buy-side clients have relied on their custodians to execute and settle the currency components of their securities trades.
In future, they won’t.
The days in which custodians could buy all the sells and sell all the buys at the prices most advantageous to their own organization are over.
Traders will be replaced by algorithms lodged on digital platforms that scour the markets for counterparties, guarantee prices at or below the mid-market rate, and then execute, time-stamp and report them in real-time.
No need for custodians to intermediate the risk any more.
So that is the end of the FX desk.
The opportunity for custodians is to build or buy a real-time FX execution and reporting platform.
Fourthly, corporate actions.
At the moment, these are the last redoubt of serious inefficiency in the global custody industry.
Extensive manual re-keying.
Multiple versions of the truth.
Minimal use of message standards.
And primitive methodologies, such as the record date.
Above all, excessive intermediation: not just by custodians and sub-custodians, but by stock exchanges, data vendors, registrars, and proxy voting agencies.
In future, issuers will key corporate actions into standard templates, verified and stored by utilities which execute the instructions of investors and then pay them, directly, in central bank money.
The opportunity for global custodians is to build or buy, and especially to operate, the utility.
Income collection will go the same way.
Issuers will pay dividends directly to investors, using lists of investors and their holdings that they receive not from registrars, and not on record date, but in real-time.
And who will need reporting of the old-fashioned kind, in which custodians fish transactions and positions out of multiple systems, and mash them clumsily together?
Distributed ledger technologies are by definition distributed databases.
These contain, as a matter of course, a record of positions and a complete history of transactions between counterparties in a particular asset.
The contents of distributed databases can be shared with regulators, as well as clients.
They also provide a convenient solution to AML and KYC demands.
But I see no reason – except inertia - why global custodians should not build and/or operate distributed reporting databases of this kind.
Seventh, securities lending and financing.
This business may well be revived by the increased collateral requirements created by the margining of derivatives.
But the work will not be done by global custodians acting as agent lenders or collateral managers.
Instead, assets will be lent and borrowed by globalised market infrastructures, which can move assets across borders at zero marginal cost.
Any global custodian with an in-house network should be able to provide just such a service.
The same is true of safekeeping.
Custodians could resign themselves to safekeeping and registration being taken over by the CSDs.
Or they could build CSDs of their own.
Properly designed, it could secure the priceless competitive advantage of settling the transactions of its clients in central bank money.
In fact, the only thing global custodians will definitely not be doing in a digital future is doing what bankers always did in the analogue age: manufacturing and advancing credit.
There will be no need for credit when every transaction settles gross, in real-time, in central bank money, 24 hours a day, 365 days a year.
It is not difficult to spot which parts of our industry are most at risk in a future such as this.
This is what the value chain looks like now.
And this is what it will look like then.
So digital technology is an existential threat.
But it is a threat that is also rich in opportunity.
And the opportunities I have outlined for custodians really reduce to just one.
Become your enemy.
Become an infrastructure.
If you think that sounds dull, think again.
Our ways of thinking about infrastructure are still trapped in the industrial age.
We think of them as public goods or natural monopolies.
We worry about who should own them.
Who should use them.
How much money they should be allowed to make.
Or how their costs should be recovered.
In other words, we see infrastructure as a supply-side problem to be solved.
In the digital age, infrastructure is a demand-side opportunity to be seized.
To be specific, an infrastructure that can store assets in digital form, and transfer value between account-holders, on a gross basis, in real-time and in central bank money, will be providing a service for which the demand is potentially unlimited.
It could, in principle, liberate a host of new services whose only obstacle to success is a lack of public trust in their ability to honour their commitments.
It could make a reality of the promise of the evangelists of the blockchain, and turn the Internet of information into the Internet of value.
It could, in principle, replace every form of banking that we know and use today.
So the opportunity is not a trivial one.
The question is whether the leadership of our industry has the imagination, and the will, to realise it.
On this point, I am less hopeful.
One reason for that is the prevailing culture of conservatism in this industry.
Nothing illustrates it quite so well as the efforts of the global custodians to keep omnibus accounts alive.
On one of the two panels at Sibos I moderated on this issue, I listed the benefits of omnibus accounts for custodian banks (I could think of four) and the disadvantages for investors (I could think of six).
When I asked the custodians on the panel to list the benefits for clients, I was greeted by a 20 second silence.
At the end of which one gamely said, “Omnibus accounts are cheaper.”
Now whatever we think of the merits and demerits of fully segregated client accounts, resistance is futile.
Regulators want them.
Clients want them.
And, yes, their abolition means the abolition of custody as a business that is recognisably engaged in old-fashioned banking.
With fully segregated accounts, custodians can no longer treat securities like cash deposits: as their property, to profit from, in whatever way they choose.
But if what I have said this morning comes to pass, that is precisely the kind of banking which is doomed anyway.
It is our past.
It is not our future.
Another reason to worry that we are not up to the scale of the opportunity is the meanness of our current ambitions.
The industry has been congratulating itself of late on shifting settlement in Europe to T+2.
Some custodians predict that the industry may one day be able to settle transactions on T+0.
Well, Amazon is open 24/7, 365 days a year, and does not seem to struggle to deliver physical units – let alone digital ones – on T+0.
In a world in which digital businesses are trading round the clock the average securities settlement system is open for six hours a day, five days a week.
Here in London, by its own estimation the most international securities market in the world, a transaction can be settled in less than one in three of the 168 hours that constitute a week.
This is absurd: physical goods are being delivered more quickly than they can be paid for.
It is an absurdity which has finally persuaded central bankers around the world to introduce real-time gross settlement of retail payments.
18 countries are now live with a domestic real-time payments (RTP) system of some sort, and another 12 countries are exploring, planning or building one.
The European Central Bank (ECB) and the Federal Reserve have both made clear that is the way they want retail payments to go.
In other words, we are moving into a world in which even the most trivial payment is going to settle gross, in real-time, around the clock, around the world.
As retail payments settle real-time gross in more and more markets, payment systems are going to link across borders.
Networks of buyers and sellers are going to go global.
That is entirely an effect of technology.
Networks based on people, and analogue technology, are hard to scale.
Networks based on digital technology can scale without limit.
So it is sobering to look at how little use our industry has made of the same opportunity.
This is a map of the opening hours of settlement systems around the world.
It looks like a Paul Klee painting, but without his respect for geometry.
Try moving cash and assets around the world, 24/7, 365 days a year, across that jigsaw puzzle.
Yet, in theory, this industry is all about automation through standardisation.
In theory, its inhabitants understand completely the value of standards.
They know that, by reducing the cost of transactions, standards increase the volume of transactions and the scale of networks.
Yet the market share of SWIFT, even in the payments industry, is probably less than 1 per cent.
In the securities industry, it is infinitesimal: it probably lies somewhere between 0.002 per cent and 0.02 per cent.
Everybody in this room should be embarrassed by that.
ISO 15022 is 20 years’ old – and many firms in this industry have still to adopt it.
As for ISO 20022, almost nobody in this industry is even interested in adoption, except the central banks and the market infrastructures.
Even the market infrastructures have freelanced so heavily with ISO 20022 that they were obliged to sign a charter at Sibos this year committing themselves to - get this - standardise the standard.
This industry is lucky that its clients are so much less demanding than retail consumers.
Which brings me to a third reason why I worry that the industry is not well-placed to respond to the opportunities created by digital technology.
I said at this event two years ago that this industry was uniquely unlucky in its institutional customers.
That they were ill-informed; distrustful; and suckers for the most expensive product on the shelf.
I stand by that verdict.
Your institutional customers are the last people to recognise and buy a higher quality, higher performance custody product at a lower price.
They do not know or care that cheaper and better solutions exist, or might exist; they are not asking for them; and they would not know how to use them even if they did.
Unfortunately, they are also the people who will ultimately determine whether or not you successfully transition your business away from an industrial age, analogue era service into a cheaper, better and more profitable digital future.
You might be willing to bet billions on a risky reinvention of your business.
But until your existing customers wake up to the possibility of something cheaper and better, that reinvention cannot possibly succeed.
As BNY Mellon explained, when it abandoned work on its CSD earlier this year, the idea was “ahead of its time” and “the pace of market change has been slower than expected.”
Unfortunately, willingness to concede defeat does not mean that the enemy goes away.
In fact, it makes your predicament worse.
It hands the initiative to the competition.
And the great advantage new entrants have over incumbents is that they have no legacy business.
Unlike you, they have no customers to lose.
They have only customers to win.
Which makes it much, much easier to build a case for investing in revolutionary strategy.
That readiness to act is in large part a function of scale.
Inevitably, even the most brilliant innovation will start with a limited number of sales.
To customers which are, by definition, less profitable.
And fewer in number.
The global custodians are also few in number.
But they are not small.
This industry is dominated by four extremely large corporations with sales running into billions of dollars.
Four out of five dollars in global custody are now controlled by just four banks.
A bank with $20 to $25 trillion in custody, and annual revenues of $10-$15 billion, is always going to find it hard to divert resources into a radical, technology-driven transition strategy.
For them, an opportunity that grows revenues by 20 per cent has to be $2-$3 billion in size.
For a $10-15 million new entrant, the equivalent sum is just $2-$3 million.
Companies that look at $2-3 billion opportunities only are programmed to kill a lot more ideas than they ever pursue.
In fact, they are bound to be inescapably, irretrievably brilliant at ignoring the great ideas their customers do not want – until their customers want them.
And by then it will be too late.
In short, global custodians are the hostages of their existing customers.
To resolve a hostage crisis without anybody getting hurt takes extremely flexible organisational structures and a special brand of astute and adventurous leadership.
At the moment, the global custodians have neither.
The banks most of you work for are organised hierarchically.
Information flows not throughout the organisation, but up and down the hierarchy.
Entire careers – many of them depressingly successful – are built on “managing upwards.”
Unfortunately, one of the few secure findings of the entire corpus of theory about innovation in business is that hierarchical organizations are exceptionally poor at it.
This is because they are exceptionally bad at processing and acting on new – and especially on uncongenial – information.
The companies that succeed in the digital age dispense with hierarchy.
They are organised as teams of equals, which coalesce spontaneously.
They value outsiders, who bring new ideas and information and practices into the organisation.
Above all, information flows not in a vertical direction between managers and workers, or between sealed groups of insiders, but freely and openly, throughout the organisation.
That does not describe many banks that I know.
Most bankers that I know work in silos that are hermetically sealed off, one from another.
Frankly, I am more likely to know the people in the adjacent silos of a bank than the people actually working in them.
As for the payments people – who understand far, far more about what digital technology is doing to the banking industry than anyone in securities services, precisely because they face off not to another business but to retail consumers – they might as well inhabit one of those parallel universes 10 to the power of 10 to the power of 118 metres away that Max Tegmark tells us about.
These silos are a source of constant frustration to clients.
Who find even service teams are split from operational teams.
That operating systems are poorly integrated.
And that nobody at a bank wants to share information with anybody else at a bank in case they lose control of the relationship.
These silos, each concentrating on its component of the securities processing chain, represent no advance at all on the pin factory described by Adam Smith in 1776.
It is the organisational technology not of the digital age, but of the industrial age.
It has never worked well.
But it has never done more damage than it is likely to do now, when the fundamental architecture of the industry is shifting on the seismic scale.
Nobody can rise to the challenge.
Because nobody is allowed or encouraged to see or think outside their own silo.
The people who should be able to, of course, are the senior management.
And how well, exactly, are they placed to steer your organisations through a period of potentially ruinous disruption?
To find out, I took a look at the most recent proxy statements of the three major pure play global custodian banks.
The people leading these institutions are all in their late fifties and early sixties.
The average age of the executive directors was 57.
Of the independent directors, 61.
After decades with the bank, they have a large emotional stake in it.
They also have a large financial stake in it.
The average value of the salary, bonus and pension of the executive directors was $17.8 million.
The average value of their unvested stock was $13.8 million a man.
If they lose their job, they collect an average of $17.2 million.
If they lose control of the company, that figure rises to $27.1 million.
These are older men –and it is mostly men – with a lot to lose.
They are, quite literally, over-invested in the status quo.
Even the so-called independent directors are independent only in the sense that a joey is independent of a kangaroo.
The fee they are paid average $112,689.
Their average holding of stock in the bank is $1.9 million.
Now, you might think people paid $10 or $12 million a year should be leading their companies into a glorious future.
But it was the past which put them there.
It is the present that keeps them there.
And none of these people is ever going to vote for an experiment which will inevitably seek to trade current margin for future volume.
If they are interested in innovation at all, it is innovation which will sustain what they are doing already.
Their response to digital technology is not to ask what new revenues it could invent.
It is to worry about what existing revenues it might destroy.
Of course, they all have R&D centres, FinTech partnerships and blockchain projects in hand.
A paper published recently by Greyspark identified 39 banks that are exploring blockchain initiatives.
But this is disruptive technology strategy 101.
You put disruptive technology in a box, where it cannot possibly attract the attention of existing customers, or threaten the existing corporate culture.
Kosta Peric, who set up Innotribe for SWIFT, used to say that the best way to protect your corporate castle is to set up a sandpit outside the walls for your children – and the friends of your children - to play in.
And the 39 banks are doing exactly that.
The trouble is, the new technologies are not promising more of the same.
Or even new and better ways of what is being done already.
They are promising a total revolution in the structure of the entire financial services industry.
It is an easy mistake to make.
After all, nobody running a FinTech start-up is going to tell a bank that it wants to put them out of business.
In my judgment, this has led to a vast under-estimation of the long term consequences of digital technology.
It is not fanciful to suppose that we are at the beginning of a revolution in our material affairs at least as great as that of the agricultural or industrial revolutions.
Its principal effect is a reduction in costs, and especially of transactions costs.
I am not so foolish as to expect men and women who make their living from adding costs to transactions to welcome this revolution.
How many of you have joked to me over the years about how much you value the persistence of inefficiencies in our industry.
And why not?
Those inefficiencies have until now determined the size and structure of the firms that compete for global custody business.
They have set the borders between what custodians - and their clients - do internally, and what they outsource to others, or offshore to other parts of the world.
They have allowed global custodians to spend a fortune patching up old systems.p>
And throwing bodies at problems those ageing systems cannot solve.
As a result, custodians enter an existential battle at huge competitive disadvantage against new entrants armed with new technology.
Not just because their technology is old, expensive and inflexible.
But because banking is the purest transaction cost of all.
Borrowing cash or securities at one price and lending them at another.
Buying a currency at one price and selling it at another.
It is hard already to measure the value which these activities add.
But when time and risk are removed from the equation it becomes impossible to describe banking as a value-adding activity at all.
The challenge facing our industry is to find a way of stepping off this trajectory to disaster.
Your senior management is not going to find it for you.
It is up to this audience – the younger people, the hungrier people, the middle management, the leaders in the trenches, not the brass hats ensconced in the chateau far behind the lines – that will have to find it.
And convince your customers to accompany you over the top, because it is in their best interests too.
That an upward trajectory exists, I am sure.
The volume of transactions made possible by digital technology is many orders of magnitude greater than the sum of all financial transactions which take place in the world today – and that volume can be expected to grow exponentially for decades to come.
The organisations which capture that growth will have mastered three principal challenges.
The first is to take command of data and use it to find counterparties, and understand the risks they represent and the prices that can be achieved, better than they have ever been understood before.
The second is the ability to build and operate infrastructures capable of clearing and settling transactions on a gross basis, in real-time, in central bank money (by which I mean money that is not initially manufactured as credit).
The third is the most difficult, but also the most important.
It is trust.
The assurance that the transaction clearing services that are offered are of the highest quality.
And that counterparties will do, always, what they say they will do.
Despite everything which banks have done to forfeit the trust of their customers, of taxpayers, and of the wider public, the evidence is that investors and consumers do not yet trust the alternatives.
The opportunity before us is to stop selling reductions in prices and start selling reductions in costs.
Costs in general, and transaction costs in particular.
The costs of finding the counterparty.
Negotiating the price.
And delivering the service against certainty of payment.
It is what Amazon does.
It is what Uber does.
And look how much trouble they are causing the incumbents.
So let me close by inviting you to ask yourself a question.
Do you want to be a Black Cab driver?
Or an Uber driver?
It is hard to accept that knowledge, painfully acquired, is obsolete.
But remember: that is why the people outside our industry see, much more clearly than anyone inside our industry, where the logical exit from our current predicament can be found.