LMJ editorial board member Joseph Paris explores the financial industries attempts at lean and shows where the real savings have been made.

Willie Sutton, a famous bank-robber during the mid-1900’s, was once asked why he robbed banks. Willie responded, “Because that’s where the money is.”

Contrast that with an ad on television several years ago, from a major American bank. The president of the bank was speaking to the camera and said, “Last year our bank processed over 10bn cheques accurately”. He then corrected himself and said, “Actually, our bank processed one check accurately last year and then repeated the process over 10 billion times”. I am sure he did not realise it at the time, but he was directly speaking of velocity of lean and the quality of six sigma.

What is remarkable is the transformation of cold hard cash having to be physically transferred when a transaction was made back in the 1900’s – to it all being bits of information exchanged from one financial institution’s computer to another’s, as it is today.

When people think of continuous improvement, they don’t think of the finance– except to believe it needs more improvement. How many of us would consider ourselves fans of our own bank with whom we do business?

But in reality, financial institutions do an excellent job when it comes to high-volume transactional activity – an area in which they have invested mightily in automation.

For instance, in the last five years, how many of us have had transactions gone awry? Posted from, or to, the wrong account? Or an occasion when the amount was wrong? What was the source – the root cause – of the error?

Whether processing a credit-card charge, posting a payment, making a deposit, or processing a cheque – all of these activities are standardised, rigid and automated. There is no room for error once the control and processes are within the system. If there is an error, it is usually at the point of entry or exit – when a human has the opportunity to engage.

As the transactions get more complex – with the number of variables and steps increasing as well as the level of risk – the processes become less standard, leading to less automation and an escalation of human involvement (labour). These circumstances also result in an increase in exceptions with the result being quality and efficiency suffers.

We can witness this with our own life experiences. The mundane transactions above occur without our giving them a second thought. But look at some of the more complex transactions, and what some of these increased complexities (and human involvement) might be;

  • Vehicle loans: requires more paperwork to be examined and accepted than everyday personal banking activities – but the process is largely standardised due to the number of transactions of this sort. Does the appraised value of the vehicle support the value of the loan being considered? Is the title clear of liens? How will the person taking out the loan pay it back?
  • Home loans: even more paperwork to be examined and accepted. There are all of the above requirements, plus the history of the property going all the way back to its discovery. Who owned it from when to when? Were all the loans throughout its history satisfied? Are there easements? Did it pass an inspection?
  • Commercial loans: now the evaluation process for the bank becomes labour intensive.  What is the viability of the company and its long-term prognosis for success?  What is the value of a custom-built facility or a specialty apparatus to the marketplace beyond the company making the acquisition? If the bank had to repossess, how much could they actually expect to get for the asset? What about the inventory? Is it really worth what the company says its worth?

Each of these transaction types are commoditised transaction. Sure, a familiarity and relationship between the parties’ helps, but the only difference from lender to lender is price and terms. The result is pressure on the sell price, and the resultant pressure on cost to deliver.

Mergers and acquisitions – where cost doesn’t matter:

Think of Facebook’s acquisition of Instagram for $1bn in April 2012. That is just the announcement of a transaction in principle, the very beginning of the merger process – two owners meeting and agreeing to a deal and the rough parameters under which it will be ultimately consummated. Since most deals get done once announced (especially when one of the parties is a publicly-traded entity), the news is usually celebrated with great fanfare.

Sometimes, the deals are not made in a friendly fashion. Sometimes, the party that is wanting to acquire will call the (publicly-traded) party which is the target of the acquisition and simply inform them of their intent. Then things can get very complicated.

What if there other serious bidders? Such was the case with RJR Nabisco, which pit a leveraged buy-out from Kohlberg Kravis Roberts (KKR) against a leveraged management buy-out. In the end, KKR won.

And are governments willing to get involved? This occurred with General Electric’s (GE) acquisition of Alstom. In this case, there was a feigned attempt from Siemen’s to get involved in the deal as a spoiler, but it was the French Government, citing national interests that was the real hurdle that needed overcoming. Once GE agreed to pay the French government, all the national interests’ issues were resolved.

Most business leaders are motivated to get deals done. After all, the executive leadership are usually going to make money when the deal is completed.

But these transactions are complex and require highly specialised experts (on both sides of the transaction) to ensure that the accuracy of the transaction and its intent is a reality. Although most deals have similar considerations, the nature of the deals is such that each one is a one-off and do not easily or obviously lend themselves to streamlining.

  • Terms of the deal:  what’s it going to take for both parties to be satisfied
  • Non-disclosure agreements: neither parties wants the private inner-workings and strategies of their organisation to be made public.
  • Employment agreements: after the deal is done, some people are going to stay and others are not. For those who are going to stay, there has to be a position and a compensation-package agreed. For those who are going to go, there has to be compensation to keep them quiet about the proprietary information they possess (and also not to go to, or become, a competitor).
  • Due diligence; each side has to ensure the other is representing itself accurately and there are no post-merger risks that are not properly accounted or undisclosed. Each of the parties must consider regulatory issues that might be relevant to the deal.
  • Definitive acquisition agreement: the formal terms and conditions of the deal must be decided and agreed.
  • Indemnifications and warranties: defines what happens if things agreed are not as they were represented. Who is responsible, under what circumstances, and to what extent?

But this does not stop after the deal is done. The companies still have to make it work. The business world is littered with the carcasses of failed mergers & acquisitions: the majority of acquisitions fail outright or fail to meet expectations.

Cost is what you pay, value is what you get.

In addition to the uniqueness and complexities involved in the transaction, there is no real motivation to cut costs. The motivation lies in accelerating the process. And if that costs more, so be it (so long as it does not negatively impact the rewards). In the example of Facebook and Instagram, the $1bn became $715m by the time the deal was done because of the drop in the price of Facebook shares (which were the bulk of the payment of the acquisition price).

For example: you are told you are going to get $1m at the end of a process. And you have the power to accelerate that process at considerable cost, but not coming out of your $1m. Remembering that time is the enemy in these deals, would you spend other people’s money to get yours?


Where do opportunities for improvement exist?

And where in the business are they likely to actually occur and have the largest impact?

We agree that what is important to mergers and acquisitions, what is value in the eyes of the customer, is velocity of throughput – and not cost. So any improvement in the processes would be measured in terms of increasing the speed of delivery with any subsequent increase in cost being (willingly) borne by one or both of the parties. But considering the deals are one-offs, that standards beyond checklists do not really exist, and the security requirements of the information being passed between the parties (and not leaking out beyond those directly involved); one can expect the scope of improvements to be incremental and not transformational.

So that leaves the emphasis of applying continuous improvement efforts for banks and financial institutions, in the activities surrounding personal banking and other high-volume activities.

Consider the cost of money (the raw materials of banking). The US Federal Reserve Bank’s funds rate is currently 0.25% – has been for some time and is projected to continue to be such for some time. In addition, the average interest paid on personal savings accounts and certificates of deposit (other sources of raw material) are around 1%.

Then consider the interest rates charged for revenue-generating transactions; a new car loan can be obtained for between 3% and 4%, and home loans for 3.1% (15-year fixed) or 4% (30-year fixed). A profit margin of 2% to 3% on deployed capital is threadbare (especially considering the indirect servicing burdens of bad-debt, infrastructure, and compliance). In fact, the only shining light for banks is in the margins realised on the interest charged on unsecured personal credit (credit cards), which can be in excess of 21%.

But there is a magnitude in volume of these types of transactions, and they all follow the same process (with very, if any, exceptions). Thus leveraging the benefits of economies of scale and lending themselves to automation (eliminating labour costs) result in considerable benefit to the bank and to the customer as well.

There is online banking, every vendor takes credit or debit cards, and there are ATMs almost everywhere. And if a customer has a problem, the banks have implemented sophisticated telephony trees, where an automated system answers the call and the caller is prompted with a series of questions intended to guide the caller to the answers they seek. Some more robust trees will also provide computer-generated answers to some routine questions such as; account balance, last charges, next payment date and amount due.

Recently, Citibank changed its credit-card fraud detection system. If a charge is suspected of being fraudulent, a text detailing the charge and asking to return a 0 if the charge is valid and a 1 if it is otherwise.

There has been an enormous amount of transformation in the banking industry in the past 20 years. Some improvements have reduced the costs to deliver financial services while simultaneously increasing the velocity of providing the services and their accuracy; such as leveraging the technologies related to the internet and telephony. And some of these same improvements have led to their being exploited for nefarious activities and a lessening of ability to engage should a need arise that is outside the system.

The challenge to the future will be: how much non-value-add-costs will banks pursue to eliminate? How much accurate and secure acceleration will customers want (but still have the ability for a human to engage when the situation is outside operational parameters)? And how much of all of this are governments willing to allow?