Thursday, June 19, 2014

A new payments system for the US

In a weird twist of fate, I was at the Federal Reserve's Payment System Improvement Town Hall in San Francisco today morning, just as Simple vastly improved the speed of our external transfers. Of course, we didn't magically make ACH any faster, just improved the end to end process to eliminate all bottle-necks. Further improvements from here are limited by the ACH system itself. If you don't know how ACH works, I wrote a blog post about it a while back. Unfortunately, at best ACH is a next day payments system for pushing funds. The US does not have a national real time retail payments system today. ACH is so outdated that the new Check clearing infrastructure that was built to implement Check21 is frequently faster.

Now there are lots of improvements that could be made to ACH, but historically they have been blocked by the large banks who make so much money from charging customers fees for pricier wire transfers. It would be almost funny if it weren't heartbreaking - less than a year after Occupy Wall Street was broken up, the big banks voted in secret to stop straightforward payment systems improvements which would have improved the lives of hundreds of millions of people.

One good thing that did come out of that was that it seems to have galvanized the Federal Reserve to launch a payments system improvement program. The Fed started a consultation process with the public, invited comments, conducted research and came up with a set of options to evaluate.

Amazingly, the option they have chosen to explore is to build a whole new payments system from the ground up!! I've been tracking this process for a while now, and I definitely didn't see this coming. Its a ballsy move by the Fed, and if they manage to pull it off, it will catapult the US back to where it was in the 1970s - a global leader in electronic payment systems ubiquity and efficiency.

Whether they pull it off is a big IF. The fundamental problem the Fed faces with most of the retail electronic payment systems (ACH, Check and Cards) is that it doesn't control them. Even with ACH and Checks where its a major operator, the Fed doesn't get to set the rules. Congress never gave them that power. So the rules are made by industry associations like NACHA, which run the systems to maximize profit for their members, not to maximize welfare for end users. To try and get around this problem, the Fed is inviting comments from all stakeholders, and trying to build consensus around this new direction. They also have a lot of soft power they can bring to bear on recalcitrant banks. But it will be a long road, and even defining the roadmap will take another year, with any form of implementation unlikely to begin before 2017.

That doesn't mean nothing will happen in the short term. Incremental improvements to the existing payment systems will happen - the Fed previewed some improvements to the NSS, and mentioned changes coming to the Check21 architecture as well. Who knows, we may even see same-day ACH become a reality the next time NACHA gets around to voting on it.

Thursday, January 2, 2014

ACH is bad, but the banks are worse

So there’s been a lot of talk about BitCoin lately, and I particularly liked this piece by Guan - BitCoin’s real value is it's payment system. It doesn't necessarily work very well as a store of value or unit of account, but the very low transaction costs and ease of signup make it attractive as a payments system. The 1% cost of some bit coin transactions is very acceptable for small payments, and its fairly typical for small payments to cost more on a % basis than large ones. That vast majority of bitcoin payments cost far less than 1%, and thats very cheap for an international payments system. 

One of the reason’s that Bitcoin has become popular is because the US payment system sucks - ACH was originally designed in the early 70s, and its way out-dated compared to most of the national payment systems across the globe. Of course most of those systems were built and deployed 20+ years after ACH, but thats kind of the point: the US payments system is outdated and creaking under the demands of today's world. Unfortunately, the big banks that monopolize banking in the US have stymied all efforts at fixing the problem

So I agree with the thrust of Guan's post, but thought it might help to clarify some things - 
  1. ACH is actually a next business day system not unlike what Guan describes in Denmark. If your bank sends an ACH payment to the Federal Reserve by midnight, the Federal Reserve will send it to the receiving bank by 6 am of the next business day. In reality, the cut-off is 2.15 am ET, so really it takes less than 4 hours. However, almost 100% of banks in the US are on batch processing systems left over from the 70s, so each of the steps of making an ACH transfer: 1)debit sender’s account; 2) send to recipient bank via Fed; 3) recipient bank credits recipient account; ends up taking 1 day, and thats what leads to the 2-3 business day time for the end to end ACH payment.
  2. The Federal Reserve charges banks $0.0025 cents per ACH transaction. Thats right, less than 1/3rd of a cent. Think about that the next time your bank tries to charge you $10 for a “Next-day payment”. That’s essentially the bank sending an ACH transaction to the Fed tonight without holding your money for a couple of days first, and charging you for the privilege.
  3. FedWire costs banks $0.138 per transaction. Think about that the next time your bank charges you $35 for a wire transfer.
  4. So if back-end transaction costs are so low, why are retail costs so high? Some markup is to be expected, but these margins are insane, especially given the massive scale of US payments systems. ACH moved $37 trillion in 2012. Because NACHA, the standards body that sets ACH standards, is controlled by the big banks, not by Congress. So essentially, the banks write the rules, and the Federal Reserve implements them. 
  5. Not all customers in the US can get a bank account easily. There are 17 million unbanked adults in the US, and 10s of millions more are "underbanked". One of the main causes is that they ran afoul of a bank some day, and got their name added to debit bureau systems like ChexSystems/Qualifile. Most banks in the country will deny an account for anybody who has a ChexSystem entry.
As even Guan realizes in his post, the payment system is one part of the problem, but the outdated technology being used by banks is another big part. There are over 7000 banks in the US, plus 1000s of credit unions, so fixing the problem of outdated batch processing technology is a herculean task. Forcing banks to adopt new systems, which might allow them to automatically post a FedWire transaction intra-day without teller involvement for example, would bankrupt a lot of the smaller banks. The big banks are even worse off technology wise; CITI has spent billions of dollars to build a new BATCH processing transaction system which they are still rolling out globally; most of the others aren't even trying to fix their systems. Why should they, when they can just pass on the costs to customers, and the regulators aren't allowing any competition?

The Reserve Bank of Australia recently moved to same-day payments in Australia, and all the banks in Australia have a timetable to move to real-time payments within 5 years. The Federal Reserve in the US doesn't even have the authority to think of such a move. So yes, retail payments systems in the US are broadly fucked for consumers. But if you're wondering why, it's because Congress and the regulators have failed consumers in the US.

Monday, March 11, 2013


At the SIBOS industry conference in Toronto in 2011, I met Taavet Hinrikus and Kristo Käärmann, the founders of Transferwise. Taavet is a startup veteran, and one of the early employees at Skype, while Kristo is an ex-consultant with experience in banking. In an eerie way, they reminded me of Josh and myself at Simple. They had a similarly big vision - they wanted to take on the impenetrable world of international money transfers and make them secure, convenient, and super low cost. I eventually ended up becoming an angel investor in their seed round, as did Roger Ehrenberg at IA Ventures, who is also an investor in Simple.

I've lived and worked in Europe, India, Japan, the Middle East, and West Africa in addition to the US, so international wire transfers are near and dear to my heart. I've never made one where I didn't feel like I was being ripped off in some way. It used to be that ignorance is bliss, but as I've learnt more and more about finance, it's all to painfully clear how the big "Money Center" banks that control the forex markets rip-off retail consumers. Not only do customers get charged insanely high fees, sometimes in the order of hundreds of dollars, they usually get ripped off on the exchange rate too.

This was brought home to me vividly when my wife had to make a transfer from Spain to London last month. We haven't had occasion to make a large transfer in the last couple of years, and it occurred to me that Transferwise might be the best way to make this transfer. But I didn't want to use them just because I'd invested in them - I wanted to use them because they were the best. So I asked my wife to look at what other services she could use.

First up was La Caixa, one of the biggest retail banks in Spain. La Caixa has been a serial innovator in retail banking for years, and is probably the safest bank in Spain. My wife literally laughed at me when I suggested she try using them to make the transfer. As excellent as their branch and ATM network is in Barcelona, nobody in their right mind would use them for international transfers. They have the highest transfer fees of almost any bank that we've used, and even charge fees for transfers within the Euro-zone, in apparent contravention of the spirit of SEPA. La Caixa was a non-starter.

La Caixa's high fees drove my wife to open an ING Direct account in Spain a few years ago. They seemed more promising - after all, simplicity and transparency are part of their brand proposition. However, I still don't know what it would have cost us to make this transfer using them. Their website has a lovely pdf outlining their differential pricing structure for transfers, which is barely intelligible, even for  a native Spanish speaker like my wife. The document also fails to mention that you can do transfers in USD - which my wife found out much later. And there was absolutely no information we could find on their website on the exchange rate that we'd get on the transfer, or what the mark-up would be on the mid-market exchange rate.

At this point it was pretty obvious that Transferwise wasn't just our best option, it might be our only reasonable one. I cheated a bit - I pinged Kristo. Our transfer was a bit unusual in that we wanted to make a USD denominated transfer to an account in the UK, but Transferwise was able to take care of it. Exchange rates and costs are displayed on their front page, and they even refunded $500 into my wife's account a few days later because the mid-market rate on the day of the transfer moved in our favor.

The problem in financial services

It never ceases to amaze me how lousy bank interfaces are in general. ING is probably the best online bank in Spain, yet it is impossible for even sophisticated users to understand how their system works. And believe me, we tried, even going to the extent of placing an international call to their CR line. Getting an idea of the exchange rate was impossible, and even understanding the services they actually offered was hard. 

The problem is that bankers think like accountants. They build systems of accounts, with complicated ways to transfer money between accounts, and then expect users not just to understand those accounts and rules, but also to optimize them. And of course there is always a fee or a commission or a skewed exchange rate ready to trip you up somewhere in the fine print if you fail to do that optimization.

The startups are coming

It warms my heart to see more financial startups taking on the incumbents, and building and shipping awesome products to customers. This is probably a subject for a longer blog post, but I realised at Money 2020 last year just how many interesting new startups there are out there: Square, Braintree, Stripe, Betterment, SoFi, Zipmark, Payperks, and OnDeck Capital come to mind, and thats just listing companies where I personally know folks. There are dozens more out there, and its clear that there will soon be hundreds. The multi-trillion $ global financial services market is going to see some disruption.

Monday, December 17, 2012

The Simple story

My speech from the FST "The Future of Banking & Financial Services" conference in Sydney on Nov 8.

Many Thanks to Marcel and the team at FST for inviting me to Sydney.

Update: I uploaded the full, hi-def video to youtube. Also, I did an interview about Simple while I was in Sydney, and you can view that at the the FST site

Saturday, March 10, 2012

The fastest growing bank in the US is Facebook

The first thing to realize is that Facebook is in fact a bank. It doesn't have a charter and its not regulated*, but it accepts deposits from consumers, and provides means for payment using those deposits. In my book, thats pretty darn close to being a bank.

Facebook Credits are really deposits
This becomes clear when you realize that you have to give Facebook money in exchange for those virtual Credits it gives you. You can use Credits for payments of course. But till you actually do that, Facebook is in effect holding that money for you. In other words your money is deposited at Facebook, and you trust Facebook to hold that money for you safely till you decide to spend it.

Facebook has to account for this money as a liability, which is what bank deposits are for banks. They even acknowledges this in their S-1 -
"Upon the initial sale of our virtual currency, we record the value purchased by a user as deferred revenue and deposits."

Deposits at Facebook are growing rapidly
And what's happening with these "deferred revenue and deposits"? They're growing rapidly by over 114% between 2010 and 2011

Of course, this is still minuscule compared to the trillion dollar deposit bases of behemoths like JPMorganChase and Bank of America. And its always easy to grow rapidly from a small base. But with Facebook set to grow its Credits volume globally, and working hard on increasing distribution, they could stay on this growth trajectory for many more years to come.

Final thoughts
By issuing its own currency, Facebook is closer to a central bank than a traditional retail bank, but that actually means it pays even less to raise deposits. And not being a regulated entity means that it doesn't have to hold reserves with the Fed or comply with any of the Federal regulations like Reg E, Reg P, etc.

It does need to comply with state money transmitter laws if it wants to enable p-2-p payments, but its already working on doing that. Given the trillion dollar size of the banking industry, even limited success could make Facebook's $100 bn IPO could end up looking cheap.

*except under the state MSB laws.

Sunday, October 16, 2011

Neither a borrower nor a lender be

"Neither a borrower nor a lender be,
For loan oft loses both itself and friend,
And borrowing dulls the edge of husbandry."
Hamlet Act 1, scene 3, 75–77

Those of you follow me on twitter know that I got a little upset earlier this week over a blog post by @. If you follow the link, you will see that the post is about an article by Kash Mansori in the New Republic.

Now, I am a big fan of Mansori, and love his blog. But stuff like this
"In particular, it misses the fact that the very design of Europe’s common currency area not only caused, but was meant to cause the eurozone’s periphery to incur large amounts of international debt. Further, there was little that the governments of those countries could do to stop it. Far from causing the crisis, the peripheral eurozone countries were up against powerful forces outside their control, forces that probably made this crisis inevitable no matter how responsibly they behaved."
is crap.

Yes, the Euro was meant to encourage capital flows from the capital-rich core to the capital-poor periphery. And yes, such capital flows frequently lead to crises when they are reversed. But thats a far cry from saying that the governments of peripheral countries are blame-free in the current Euro crisis. Essentially Mansori claims that the peripheral governments got a once-in-a lifetime windfall, and couldn't help but squander it. What Mansori misses is that the sovereign governments of those peripheral countries had a choice on what they could do with those capital flows. Instead of productively investing their capital bounty, the countries of the periphery squandered it on a consumption binge that lead directly to the problems those countries face today.

Lets look at the countries that Mansori mentions in his article:
  • Greece: Greece has had a history of being a serial defaulter ever since it became independent from the Ottoman empire in the early 19th century. Its no surprise that faced with a massive influx of capital, successive Greek governments choose to squander it on a massively bloated public administration, pension promises, and whatever else they needed to do to get themselves re-elected, including fudging official statistics. The Greek government choose to run a fiscal deficit in good times; its no wonder that in bad times they find themselves in a crisis.
  • Spain: Spain actually ran a small budget surplus during the boom, and even today its overall government debt-to-GDP ratio is pretty low. But Spain had a massive housing boom, among the largest in the developing world. The Spanish Central Bank tried its darndest to ensure that local banks followed safe lending practices, but they inevitably ended up with massive exposures to local households and builders. 
  • But the real reason that investors have turned away from Spain is because of the parlous state of its economy. Unemployment currently runs at over 21%, and even the meager growth forecast of 1.3% for 2011 looks optimistic. You don't have to look far to find the reasons - permanent employees get 6 weeks of severance for every year they have worked. Once you've worked at a company for a few years, you can pretty much stop working, because firing you is now impossible. Younger employees are limited to temporary contracts of 1 year or less. They are always the first to be fired, and employees have no incentives to invest in training them. As if that wasn't bad enough, bureaucratic red-tape means that incumbents in many industries are entrenched and entrepreneurship practically non-existent. Even MBA students from the best universities dream of getting a bureaucratic job - those are the only good jobs still available to the young. For most of the last decade, wages grew rapidly and productivity growth was flat to negative. 

    Massive reform will be needed before Spain can return to growth, and that sort of reform looks politically impossible at the moment. With 9% fiscal deficit, no way to devalue the currency, and no hope of growth, even Spain's relatively light debt burden begins to look heavy. Small wonder that capital (and mobile labour) is fleeing the country. 
  • Ireland: In some ways, Ireland is the poster child for Mansori's argument. It has one of the most flexible and business friendly economies in Europe, with famously low corporate taxes. It enjoyed rapid productivity growth and attracted large numbers of skilled immigrants for most of the period leading up to the crisis. True, it had a huge real-estate bubble, but arguably the government could do little about it in such a lightly regulated economy.
  • In fact, all of Ireland's problems stem from one decision made in late 2008 - the decision to guarantee the debts of all the Irish banks. These banks had borrowed heavily in the global markets to finance a wave of expansion as well as the domestic housing bubble. Banks like Anglo-Irish were effectively bankrupt in 2008, and the government stepped in to guarantee all their debts. In doing so, they instituted a massive transfer of wealth from Irish taxpayers to foreign creditors. The joke at Davos in early 2009 went something like this "What's the difference between Iceland and Ireland? One letter and about six months". It was to avoid precisely that situation that the Irish government guaranteed bank debt. But today, the Icelanders look like the smart ones - their crisis cost them 13% of GDP, while re-capitalizing its banks will cost Ireland 36% of GDP.  With Irish debt-to-GDP set to hit 123% by 2014, investors are rightly concerned.
Ultimately, a freely functioning market exerts control on both borrowers and lenders through the fear of default. Borrowers fearful of being cut-off from credit invest their borrowings wisely instead of spending them, and lenders fearful of losing their money scrutinize borrowers and monitor their behavior. When the Irish bailed out the (European) creditors of their banks, they undermined this system, and it backfired on them spectacularly. In essence, the Irish wasted a decade and a half of good governance in a show of pride over the strength of their banking system. Germany and the Nordic countries who are being pressurized to bail out the rest of Europe should take note. 

Thursday, August 18, 2011

Lessons from the Long Depression

The below chart caught my eye earlier today, and I tweeted about it. The chart itself is a bit hard to read, but essentially its telling us that long term interest rates haven't been this low since the 1960s. The chart ends on June 30th - long term interest rates have dropped sharply since then, and the 30 yr treasury is now below 3.5%, and the 20 yr is at 3%. The last time long term rates were this low and falling, was at the start of the Great Depression.

A lot of people have written about how the 2008 financial crisis almost took us into a second Great Depression, but I think its pretty clear now that we at least dodged that bullet. It's the other time that long term rates dropped this far and kept dropping that interests me- the Long Depression from 1873 - 1896.

The Long Depression

The world of the 1880s was very different from today of course. There was no Federal Reserve, the US was on the gold standard, and the west was still being won. There was no Social Security, Medicare, or safety net of any sort really. In fact, compared to the more developed economies of Western Europe, the US was pretty much an emerging market.

And yet there are similarities. The Long Depression followed a period of debt build-up driven first by government borrowing (for the Civil War) followed by a (railway) construction boom. Then as now, inflation was never a concern (the Coinage Act of 1873 moved the US to the gold standard) - rather creating inflation was a problem. The crisis of 1873 also started in Europe (in Vienna) and then spread to the US. While bank failures spiked in the US, they were more a result of the crisis than the cause.

Another Long Depression??

The bailout of the banking system in the US in 2008, and global monetary and fiscal stimulus in 2009, at least prevented a Great Depression style economic collapse. The emerging markets of today have recovered and are roaring again. But having avoided another Great Depression, are we slipping into another Long Depression?

The main difference - while the Great Depressions was a steep sharp collapse followed by a relatively steady recovery, the Long Depression period from 1873-1896 was characterized by a sustained slowdown in overall growth. If the Great Depression was V shaped*, the Long Depression was L shaped. Recession was an ever-present specter - even after the recession ended in 1879, the US economy endured many periods of contraction. It was not till long terms rates began increasing around 1900 that the US entered another period of sustained growth. Economic performance was even worse measured in per capita terms, given the huge immigration boom of the late 19th century.

Lesson for us

Part of the reason for that huge immigration boom was that Europe was way more f**ked than the US. Italy, Austria-Hungary, and France were the worst hit, while Germany, the UK, and Russia did relatively better.

The most common explanation for the Long Depression is that it was caused by a long period of deleveraging, made worse by a slavish adherence to the gold standard. The US had better demographics and cheaper land, and combined with a slightly more flexible monetary policy, this allowed it to perform better than most parts of Europe. Faced with a similar but even worse crisis in the Great Depression, western governments eventually gave up the gold standard.

For Europe today, the hard money policies of the ECB are the equivalent of the gold standard. Europe is effectively on a Deutschemark standard. If the Long Depression is any guide, than the Euro can survive this episode, but it could very well mean a generation of lost growth for many parts of Europe.

The Fed's willingness to print money, and the US' old advantages of demographics and land, should ensure that the US will continue to outperform Europe. But sustained growth of the pre-crisis kinds may not be achievable, especially as it looks like the Fed is close to being tapped out. The US may even have to deal with another influx of immigrants from Europe.

A Third Industrial Revolution

The late 19th century was also a time of great technological advancement known as the Second Industrial Revolution. Steam shipping, railroads, the growth of the telegraph network and the opening of the Suez Canal led to an explosion in trade and eventually created the conditions for a global recovery around the turn of the 20th century. The countries that tapped into that burst of entrepreneurial energy - Britain, the US, and Germany to a certain extent, were also the ones that came out of the Long Depression in the best shape.

The Internet and the myriad innovations it has spawned are creating a Third Industrial Revolution today. The concept has been much discussed, but nobody can predict today what innovations will drive growth for the next few decades. The only thing that governments can do is foster an environment that reduces red-tape, and makes labor and capital as easily accessible and accurately priced** as possible for the entrepreneurs. The countries that do that well (my money is on the old trio- the US, UK, and Germany) will probably outperform again.

And as for individuals, don't curse the darkness, light a candle.

* more W shaped then V really.
** we learned that lesson from the dot com bubble I hope

Saturday, March 12, 2011

The End of Interchange?

Last week, the Federal Reserve Board requested comment on proposed regulation that would “establish debit card interchange fee standards and prohibit network exclusivity arrangements and routing restrictions.” The proposal is an outcome of the Dodd-Frank Act, which stipulated that interchange fees be “reasonable and proportional” to the cost incurred by debit card issuers.

The Fed proposed two schemes for setting interchange rates, but effectively capped interchange at 12 cents per transaction. Today net debit interchange is around 40-45 cents. In effect, the Fed is planning to reduce interchange by at least 70%.

The cap was lower than many people expected and came as a surprise for the industry. In response to the announcement, shares of Visa and Mastercard plunged more than 12%. While this may hurt our own bottom line, we support the direction of the Fed’s proposal.

Because most people who regularly use debit cards are unaware of the associated interchange fees, it’s difficult to decipher what this proposal means for consumers. Doing so requires an understanding of the history of interchange.

History of Interchange
In the 1950’s, when credit cards were introduced, most people bought with either cash or checks. Regular customers with good credit would frequently have accounts at stores, and would get a bill at the end of the month for all their purchases. Credit cards allowed people to buy stuff from more stores without having to carry cash around. They were also safer to user and more convenient than carrying checkbooks around, so consumers loved them. Merchants liked them too, since they brought in new customers who merchants otherwise wouldn’t have been able to extend credit too.

Of course, building this system and marketing it to customers was extremely expensive in an era when calculators were just beginning to catch on. So the credit card companies charged merchants a fee for accepting payment via a credit card: instead of getting 100% of the value of a credit card transaction, the merchant would get 95%-97%. The credit card companies also charged customers interest if they didn’t pay on time, and the combination of interest and merchant discounts was enough to make the business profitable.

Over time, the industry grew and evolved. Banks got into the game and begin to issue credit cards through associations that eventually became Visa and Mastercard. Credit card acceptance usage and acceptance exploded in the 1960s and 1970s to the point where they were ubiquitous. The card associations were owned by the banks, and paid them most of the discount fees that they earned from the merchants, which became known as an interchange fee.

Success of Debit
The success of credit cards demonstrated the need for a safe and convenient alternative to cash. But it wasn’t until the emergence of debit cards when card based payment became commonplace.

Debit cards began as a byproduct of ATM cards in the 1980’s. Because ATM usage was immediately popular, banks began engineering a method for customers to use ATM cards as an alternative to credit cards. Because few merchants had terminals that could accept PIN based ATM cards, debit cards were designed to work with merchant’s existing credit card terminals, and to route transactions over the existing credit card network.

Today, debit cards transactions are still routed through two separate payment networks---signature debit, where customers provide their signature and the transaction is routed over the credit network, and PIN debit where the customer enters their ATM PIN code and the transaction is routed over an ATM network. Regardless of how the transaction is routed to the card issuing bank, the money always comes out of a user’s checking account.

Debit cards also had interchange fees associated with them, just like credit cards. Initially, the interchange fee that merchants paid on debit card transactions was much lower than what they paid on credit card transactions. While cash and checks didn’t have any fees associated with them, they had hidden costs. They had to be stored under lock and key to prevent them from being stolen, somebody had to deposit them at the bank, and there was always the risk of getting a bad check or counterfeit cash.

Compared to the alternatives, debit cards were relatively cheap for merchants to accept. Debit cards were also available to everybody with a bank account, regardless of their credit history. Furthermore, banks loved having an additional source of revenue from card fees, especially from customers who either wouldn’t or couldn’t get a credit card. So everybody loved debit cards. Not surprisingly, debit card usage grew rapidly throughout the 1990’s and 2000’s. A whole segment of the industry, prepaid cards, came into existence to provide debit cards to people who didn’t even have bank accounts. It wasn’t long before most Americans had a debit card in their pocket. In 2009, debit cards overtook both credit cards and checks to become the most widely used payment mechanism in the US.

However, as debit card usage grew, the industry structure changed. Mastercard and Visa realized that debit was a huge growth opportunity and began to compete aggressively for customers. Most people already had checking accounts by this time, so Visa and Mastercard were competing for the right to be the logo on a bank’s debit card. The quickest way to win a bank over was to offer it more interchange revenue from merchants. So Visa and Mastercard consistently raised interchange fees charged to merchants on debit card transactions.

Merchants didn’t like these ever increasing fees, of course. Because most merchants had a terminal that can accept a debit or credit card with the Visa or Mastercard logo, the networks had enormous market power. The merchants had no ability to negotiate with Visa and Mastercard, who up until recently were owned by the banks. Visa and Mastercard logically deployed their market power in the interests of their shareholders and customers, the banks.

These changes were hidden from cardholders. Marketing and rewards programs persuaded customers to use cards rather than cash. The cost was paid by merchants in the form of increasing interchange fees. Even when the cost of processing declined---as payment processing switched from paper to electronic---interchange fees continued to rise.

Congress to the rescue
The system as it stands today is clearly broken. Banks benefit as the cost of processing decreases and interchange fees increase. With their tremendous market power and lobbying capabilities, banks ensured that nothing much could be done about it for a long time. In this, as in many other things, the financial crisis changed everything.

Last summer, Senator Dick Durbin from Illinois attached an amendment to the Dodd-Frank Act that required the Federal Reserve Board to regulate interchange fees. As soon as the banks realized what had happened, they lobbied furiously to have it removed, but the financial crisis had weakened their position. With minor exceptions for small banks, government benefits cards, and prepaid cards, the amendment became part of the final act.

The Durbin amendment is clearly a flawed piece of legislation. For one thing, it completely ignores credit interchange. Furthermore, the precedent of the Fed setting prices in a private market is troubling, and it’s not something the Fed likes doing. The Fed also doesn’t have complete freedom to regulate interchange, but must look only at the incremental costs of processing transactions. There is a serious risk that over-regulation will kill the flexibility and innovation that the banking world badly needs.

How this impacts consumers
The Board’s proposed regulation will change the debit card industry. Interchange fees will no longer be a significant money-spinner for large banks. Because most rewards programs are funded by interchange revenue, they might disappear entirely on debit cards.

To compensate, the large banks will be forced to charge higher fees to some customers, especially those who only open up a low balance checking account.

Merchants should see big cost savings, running into the billions of dollars. But if the experience of Europe or Australia is any guide, little if any of this will be passed on to customers.

Friday, March 11, 2011

The rise of Credits

Facebook recently announced that the only payments mechanism on the Facebook platform will be Facebook Credits starting from July 1. As with anything related with Facebook, hard numbers are difficult to come by, but by all accounts Facebook Credits is growing rapidly, and forcing through this change will only help that.

Credits could be revolutionary

Credits are essentially a private currency. The US has a long history of private currencies. Private currencies usually function as replacements for the US dollar in a particular geography, where they can help to keep money in the community and drive business to local stores. In that way. The fact that Credits is "virtual" doesn't change its fundamental nature; it's a private currency usable within a specific community.

What makes Credits exciting is the size and reach of that community. At 500 million users and growing, the Facebook userbase is already larger than the population of the US. In fact, Facebook's population is exceeded only by that of India and China. Comparing population growth in China with Facebook's growth, Credits could easily become the most widely accessible currency in the world by the end of this decade.

Of course, the fundamental purpose of any currency is to act as a means of exchange*, and Credits won't really be useful until more people, especially merchants, accept it as a form of payment. That's where the Facebook API comes in. For the thousands of app developers on Facebook, Credits is now the only way to get paid for their efforts. This guarantees a rapidly growing base of merchant acceptance for Credits. Given how addictive many of the games built on Facebook's API have proven to be, this move by Facebook guarantees a steadily growing demand for Credits.

Of course, comparison's with earlier virtual currencies such as World of Warcraft's Gold, and Second Life's Linden dollars is inevitable. And in the early stages, Facebook Credits certainly resembled these earlier virtual currencies. But the twin factors of Facebook's massive userbase and rapidly growing acceptance through the Facebook API guarantee that Credits will be much more revolutionary than anything that has come before.

Facebook Credits is not PayPal

Another comparison that I've heard made is between Facebook Credits and Paypal. Paypal was a revolutionary p-2-p transfer mechanism when it launched 12 years ago. Since then it has evolved into a money transfer behemoth, processing over $60 billion in payments volume in 2008. Its primary revenue stream is from helping merchants process payments online. Paypal is definitely the premier online money transfer mechanism today, but it's a very different from Credits in many ways -

  1. Despite being seven years older, Paypal doesn't have nearly as many users as Facebook. The size and engagement of Facebook's userbase gives Credits a gigantic advantage.
  2. The Facebook API is another huge advantage; Paypal's X API is no comparison.
  3. Paypal doesn't have a virtual currency. Any money in your Paypal account is denominated in dollars. Credits is a virtual currency, and because of its convertibility into multiple currencies, it's like a private version of the IMF's SDR.

Credits needs work

Credits still has a long way to go to reach its full potential. Facebook needs to do a few things -

  1. Mandate Credits for everything on Facebook. Not just third party apps, but accept Credits for stuff like buying ads.
  2. Make Credits freely transferrable between users. This will immediately increase the utility of Credits for users; Credits will immediately become a money transfer mechanism reaching 500 million people at one stroke.
  3. Reduce the effective interchange rate, currently an eye-watering 30%. There is no way that Credits will be usable as a real world payment mechanism until this comes down to <5%.
  4. Get ahead of the regulatory curve. In the US, apply for money transmitter licenses in the required 48 states, and in other countries get a banking license or partner with an existing in-country bank. Paypal did it; so can Facebook.

Facebook would be smart to do all the above in a carefully phased manner. Start by pushing the use of Credits on the platform and allowing users to transfer Credits between themselves at par. This will create an incentive to hold Credits and use them as a payment mechanism rather than convert them immediately into dollars**. Then start reducing the interchange rate to convert Credits into real currencies. Do this at a steady rate, say 1% a month. Again, people will be incentivized to hold on to Credits for as long as possible, rather than convert into other currencies. All of this should be sequenced with the appropriate state, federal, and international licensing. Done carefully, this should ensure that any revenue lost from reduced interchange is more than made up through massively increased volume and seigniorage.

A Facebook Bank anybody?

Building a currency** has always been a Catch-22. People don't want to use it until lots of merchant's accept it, and merchants' don't want to accept it till lots of people use it. Unable to force usage by writing a law, private currencies have always solved this problem by persuading a small group of merchants and users to use it in a particular geography.

The flip-side of the Catch-22 is strong network effects***. Once enough merchants and users start using it, there is tremendous value for additional users and merchants to join in, and the system can grow very rapidly. The only limiting factor becomes the total population size and the scaling speed of the underlying technology. If Facebook plays its cards right, Credits will grow exponentially; much faster than Paypal ever did. I even have a bet going on LongBets about it.

Essentially, Facebook could become a private central bank, issuing a widely distributed and accepted currency. They could build a leveraged balance sheet, just like any other central bank, and with a degree of control and flexibility over their 'economy' that central banks can only dream of. Of course, they would have to develop expertise in managing a multi-currency balance sheet and hedging risks, but those skills can be imported into the organization fairly easily.

Another intriguing possibility that the combination of a virtual currency opens up is a virtual bank. A bank within Facebook, that takes deposits in Credits, and lends to anybody who needs it****. So will we see a Facebank sometime this decade?

* and a store of value, and a unit of account, and a standard of deferred payment.

** or Ringgit, or Euros, or whatever

*** same goes for building a payments network

**** initially, probably startup gold-farms and poker players looking for an ante

Monday, November 8, 2010

Why the Fed pays interest on reserves

I was at the Shadow Fed meeting in early October, and there was a lot of debate about the merits of Quantitative easing, which the Fed has since announced. One question from the audience stuck in my mind - "Why doesn't the Fed stop paying interest on reserves, and give the banks some incentive to lend that money?".

Fed Open Market Operations
Historically, the Fed has maintained its target overnight rate by open market operations. When the Fed funds rate drifts above target, the Fed injects money into the market by buying loans, and when it drops below target, the Fed sells those same loans to bring it back up to target*. When the Fed injects money, it gets converted into bank reserves, which the banks then lend out into the market, thus increasing supply and reducing rates. This essentially requires the Fed to use its balance sheet of securities, and target the level of bank reserves in order to maintain a given level of interest rates.

The Role of IOR
Paying interest on reserves allows the Federal reserve to separate interest rate policy from bank reserve policy. As Marvin Goodfriend puts it .

"Why might a central bank value the latitude to pursue separate interest rate and bank reserves policies? Interest rate policy could continue to be utilized to maintain overall macroeconomic stability. Bank reserves policy could then address financial market objectives. In the long run, a central bank could pursue an objective for bank reserves to optimize the broad liquidity services provided by the floating stock of government debt. In the short run, for example, a central bank could increase bank reserves in response to a negative shock to broad liquidity in banking or securities markets or an increase in the external finance premium that elevated spreads in credit markets. The increase in bank reserves would help to stabilize financial markets by offsetting the temporary reduction in the private supply of broad liquidity. The latitude to pursue bank reserves policy and interest rate policy separately would be useful to the extent that shocks in financial markets and the macroeconomy are somewhat independent of each other. "

Essentially, paying interest on reserves allows the Fed to put a floor under overnight interest rates. In response to a "shock to financial markets", i.e., a financial crisis, the Fed can stabilize the market by flooding it with reserves secure in the knowledge that overnight rates will be maintained by the interest rate paid on reserves. The bank can directly control interest rates, without worrying about its balance sheet or the impact of injecting more reserves into the market.

Timing of introduction
Its no surprise that the Fed introduced the policy of paying interest on reserves in October 2008. One result of the financial crisis was that the demand for reserves skyrocketed, and the Fed had to flood the market with reserves in order to stabilize it. This made it increasingly difficult to keep the actual overnight rate anywhere near the target rate as is clear from the graph below
In those crazy times, IOR helped the NY Fed to put a floor below interest rates and stabilize the market, just as it was supposed to.

Why continue paying IOR?
With interest on reserves being paid at 0.25%, IOR currently serves as nothing more than a floor to keep overnight rates somewhere in the Fed's 0 - 0.25% range. Despite what some might say, its hard to believe that cutting IOR to o.15% or 0% would do much to force banks to lend out their excess reserves. And without any stabilizing influence on the market, you might start seeing destabilizing swings of the October 2008 variety again.

A side-effect of IOR is that it allows the Fed to manage interest rates without using its balance sheet. With open market operations, the Fed has to sell securities to raise interest rates. This raises two problems: firstly, the Fed may not have enough securities** to raise the rate as high or as quickly as it wants to; and secondly, the Fed may be forced to take losses on the securities it sells if it has to sell them at a price below what it paid.

The second problem is likely to be the biggest factor in the Fed's continued payment of IOR. Despite its ongoing efforts to flood the market with reserves, the Fed will soon have to start planning its exit from the current low interest rate environment. If it used the securities on the balance sheet to do this, it would have to sell them at a loss, and deplete its capital.

So the final reason for the Fed to maintain an interest on reserve policy is independence from Congress. Essentially, paying interest allows the Fed to issue its own riskless overnight debt, instead of having to rely on riskless debt created by the Treasury. When the time comes to raise interest rates, the Fed can do so using its own debt, and not beholden to the Treasury or to Congress, either for debt issuance, or for additional capital.

* The NY Fed actually uses repos or reverse repos to do this.
** Or enough of the right kind and duration.

Friday, March 26, 2010

Banking regulation and wages

It looks like the furore over outsize pay in banking is dying down. Kenneth Feinberg's final report on the pay practices identified $1.6 billion of "ill-advised" pay among banks that took TARP funds. Yet it named no names, and insisted that the pay was all legal.

The transatlantic gap on pay looks set to widen. While the European commission is requiring that 40-60% of bonuses be deferred for three to five years and half the upfront bonus be paid in shares, pretty much all that the Dodd- Frank bill does is require that share-holders have a non-binding vote on executive pay. The logic behind Dodd-Frank's "say-on-pay" provision is unclear; after all equity holders generally stand to gain as much as executives when the risks pay-off, and it's society that ends up paying when the risks don't.

Interestingly, a study by Thomas Philippon and Ariell Reshef shows that there is a strong correlation between regulation of finance and the wages of financiers. The wave of regulation enacted in the Great Depression not only prevented a systemic crisis for over 70 years, it also had the secondary effect of reducing the wages of bankers to roughly the same level as other non-farm workers. As the chart below illustrates, the deregulation over the last 30 years that arguably contributed to the present financial crisis, also led to a dramatic increase in banker's wages.

So will the Dodd-Frank bill lead to the same sort of precipitous drop in banker's wages as the Glass-Steagall bill in 1933? Too early to tell, but as I noted in my post on BankSimple, it doesn't looks promising.

Sunday, February 7, 2010

I love Felix Salmon's blog, especially his take on the unburst property bubble, but I have to take issue with his post on the recent wave of eurozone worries and their link to market declines.

The link between newsflows and market perception is a lot more complex than simple cause and effect, and has been for a long time. The best way I've found to think about this is that markets accurately reflect the conventional wisdom in asset pricing. As Keynes put it over 70 years ago

"It might have been supposed that competition between expert professionals, possessing judgment and knowledge beyond that of the average private investor, would correct the vagaries of the ignorant individual left to himself. It happens, however, that the energies and skill of the professional investor and speculator are mainly occupied otherwise. For most of these persons are, in fact, largely concerned, not with making superior long-term forecasts of the probable yield of an investment over its whole life, but with foreseeing changes in the conventional basis of valuation a short time ahead of the general public. They are concerned, not with what an investment is really worth to a man who buys it “for keeps”, but with what the market will value it at, under the influence of mass psychology, three months or a year hence. Moreover, this behaviour is not the outcome of a wrong-headed propensity. It is an inevitable result of an investment market organised along the lines described. For it is not sensible to pay 25 for an investment of which you believe the prospective yield to justify a value of 30, if you also believe that the market will value it at 20 three months hence."

If that's the case, then the best way to understand the market declines in the last few days, is that conventional wisdom has adjusted its view on what the stock markets in general are worth. What could have driven this adjustment? Felix is quite right in saying that the fiscal status of the Eurozone countries has not changed. But information about it has become much more readily available, both in Greece where the new government in October revealed that the fiscal deficit was twice as wide as forecast, and in Spain where the government recently revealed that the fiscal deficit for 2009 was two percentage points higher than forecast. And while the MSCI exUK index is down 6.9 percent for the year, the Spanish Ibex is down over 20%, and Athens is down 16%.

Both Greece and Spain have much further to go. Greece has a gaping 13% hole in its budget, with debt pushing 110% of GDP. Spain is in a better position on the debt side, but international investors are right to worry about a country with unemployment at 20%, rigid labour markets, declining productivity, and a government that has little control of its regions and has stuck its head in the sand until recently. Belt-tightening by the Spanish government in 2010 will be at the worst possible time, a it coincides with the withdrawal of ECB monetary stimulus and the economy remain the last G-20 country still stuck in recession.

“For the past 10 years we have had an abstract, theoretical debate about whether a monetary union can work without a fiscal union,” says Marco Annunziata, chief economist at Unicredit, based in Milan. “Now we have had it tested – and have found out that that it does not.”

Once you start going down that road, there are a lot of larger European countries that start becoming suspect - Italy, the UK, even France at some level. All three have much higher levels of debt than Spain, weak economic prospects in 2010, and the UK has a fiscal deficit even wider than Spain's as well as a national election in 2010. After the "98 Asian crisis" are we going to end up seeing a "2010 European crisis" ?

Tuesday, January 19, 2010

I realized the other day that the consumer banking system in the US is highly regressive. Regardless of which business line you look at, banks across the globe target richer people. Whether its Ultra High Net Worth Individuals (assets >$30 million) in private banking, affluent customers (income >$100k)in retail banking, or mid-sized enterprises (revenue > $10 million) in small business banking, size really does matter for a bank. But while these customers generate high profits for the bank, the actual profit margin that the bank makes on them is usually much lower.

For example, a low-income customer may have $1,000 in her checking account, on which the bank pays her no interest but makes 3% a year. A richer customer may have $10,000 in his account, but will probably keep much of this in a savings account where he gets paid interest. The banks net interest margin, the difference between interest paid vs. interest earned, may be no more than 1.5% on this savings account. Of course, 1.5% of $10k > 3% of $1k, and that is why the bank prefers the richer customer.

However banks in the US are especially perverse in how they screw over poorer customers. Banks in the US make a large chunk of their revenue from charging overdraft, bounced check, late payment, and a host of other fees that the average consumers know nothing about until they see it on a bank account. Overdraft fees alone have been estimated at $38 billion in 2008. And these fees are especially regressive, since they are paid disproportionally by lower-income customers:
- lower income customers usually have lower balances and more variable incomes, making them more vulnerable to fees
- their knowledge of the banking system is limited, so they just don't understand things like a bank check taking 5 days to clear
- they have a very low success rate in calling a bank and getting their fees and charges reversed

Once you include the fees and charges, the bank may be making an effective profit margin of 5% or even 10% on that customer with a $1,000 balance, vs 1.5%-2% on the $10k customer. And this is before interchange or bill payment revenue (a topic for another day). And lets not even talk about the 40 million+ people who don't even have a bank account and relay on things like loan sharks and payday lenders.

Now this is also true for a lot of other products and services. For example, lower income customers usually end up paying more for the same groceries than a high-income customer. But in exchange for that higher price, they usually get some value in the form of smaller package sizes, distribution via a store close to their home, etc. I haven't heard of any decent-sized bank in the US tailoring a product for lower-income customers (except to try and charge them higher fees), and banks in the US do the best they can to open branches as far away from lower-income customers as they can.

Is any of this going to change anytime soon? Startups like Wonga are chipping away at the payday world, but it really needs somebody like Tesco to go after it. Sadly, all the action seems to be in the UK for now.

Sunday, December 6, 2009

So why high velocity? I realize that a lot of my apparently random interests: economics, martial arts, business, etc.; are united by a common thread. The quest not just for speed, but for speed with direction.