Atomic Energy

In an earlier post we discussed what are the reasons that a firm, at least theoretically, decides to merge with or acquire another firm. These basically fell into two categories, horizontal integration and vertical integration. And yet, we know that a huge number, 70% to 80% or more, depending on who you ask, fail, most miserably. A question that recurs regularly is why these events fail, and, by extension, what can be done to reduce the failure?

In the closing of the previous post, we looked at the very basic mathematics. For an M&A transaction to not be considered a failure – which is distinct from achieving the publicly stated goals, something that is much more difficult – the total return from the transaction, the value of the combined entities, must be at least 1 penny (or pence, or ruble, or yen, etc.) more than the amount paid plus transaction costs. Of course, almost every acquisition involves payment of more than the value of the firm on its current trajectory, so there is the added burden. Thus, the total amount gained must be at least one penny more than:

1. The current value of the firm, which is normally the market capitalization for a publicly held company, or some discounted cash flow (DCF) or similar valuation for a privately held firm. For example, as of close of trading on Friday, September 12, Merrill Lynch was valued at $17.05 per share with 1.53BN shares outstanding gives a value of $26.1BN; plus
2. The premium paid for the firm, i.e. the amount that was paid above and beyond the current value. Bank of America agreed to buy Merrill Lynch this past Sunday, Sept 14, 2008, for $44BN, a premium of about $18BN; plus
3. The costs of the transaction. These fall into several categories, and include investment bankers fees, often around 7% of the value of the deal, lawyers fees, and other related fees; plus
4. The costs of integration. These are the costs of actually putting the two businesses together. These involve everything from managing people and conducting hiring or layoffs, through systems integration and books merging, to joining sales forces and merging headquarters. 

Offsetting all of these, of course, is the value gained. This value is really composed of two very simple components:

1. The current value of the firm, which should be fairly identical to #1 above; plus
2. The integration value, i.e. the horizontal and/or vertical value added discussed in the previous article.

Looking at the above, the basic reasons why so many M&A fail to achieve real value and usually lead to a loss is as follows. The first three costs and the first value are fairly fixed and straightforward, known right from the beginning of the transaction. The real killers in the deal are always the costs of integration and the integration value. Inevitably, in every failed transaction, either the integration costs are much higher than expected or the value of integration is much lower than expected, or both. 

Costs of Integration

Why do the costs of integration end up being so much higher? To be fair, I cannot think of a single transaction I have seen, read about or been involved with that the costs of integration estimated upfront are even close to the final costs of integration. Most firms that have some form of integration have some element of integration that is literally left hanging for many years thereafter. There are a lot of reasons for this cost underestimate: 

• Humanity: People are human and make mistakes. They make the same mistakes again and again. What holds true for renovating a kitchen or building a house – the final cost is always much higher – holds true for merging firms. 
• Secrecy: Very often, these transactions are discussed in secret, due to fears of market manipulation, pressure to complete or undo the deal, or regulatory requirements. This secrecy causes many employees – those who daily deal in facilities, sales, technology or other areas that would need to be integrated – to be unaware of the deal and unable to give their input. The same management that seeks this input as critical before engaging on expansion into a new line of business avoids it before engaging in a merger of acquisition.
• Consultants: Coming from a consultant, this may sound self-defeating, but consultants are often a source of the problem. Many of the large consulting firms bill out untold millions in consulting hours to plan and assist in the integration. First, in order to get the business, they will often underbid, expecting to make up in “additional hours” or “change requests” to the contract. These additional hours come through as additional transaction costs. Second, if they feel the need to stick to the bid, they will do less work than management expected, leaving the clean-up to the in-house staff, creating additional hidden costs. Third, the large consulting firms have an unenviable track record of putting large numbers of young, just-out-of-college-or-business-school associates on the job in order to bill lots of hours at a higher profit margin. These associates, quite frankly, do not know how to manage the integration, let alone the people involved. Fourth, to be brutally honest, many of the partners at these firms do not have a clue how to piece together the elements and costs of integration. Many have climbed the consulting ladder by billing our more hours, but have never had to actually live through and manage an integration from the corporate side, i.e. pay the price for failure and receive the rewards for success. I have seen consultants from PriceWaterhouseCoopers, Ernst & Young, Accenture and all of the other firms destroy the goodwill within a firm for the transaction by their heavy-handed behaviour and absolute ineptness. Even the high end firms such as Boston Consulting and McKinsey have shown similar lack of true understanding of the real world. 
• Hubris: It had to come up eventually. Executives often have inflated views of their own abilities to manage the integration. Further, these numbers often come from the CEO, rather than the COO and line managers who really know what it takes. 
• Incentives: CEO pay is often tied not to profit margins or comparable margins, but absolute revenues, profit or comparables. Thus, the CEO of a $1BN revenue $200MM profit company will make a lot more money as the CEO of a $2BN revenue $300MM profit company, even though the margins have now shrunk from 20% to 15%. Many of these contracts are not available to the public – and kudos to John Mack of Morgan Stanley who instituted high transparency and close tie of compensation to company performance, as well as full annual director elections, after his successful coup against Phil Purcell in June 2005.

Value of Integration

From the value of integration side, similar factors come into play. Executives consistently overestimate the value the integration will bring to the merged entity. The reasons are actually quite similar to the reasons costs are underestimated.

• Humanity: People make mistakes, and executives are no different. People overestimate how big a market will be, how much value cross-selling or how much cost-saving will bring. Using the prior analogy, people invariable underestimate the cost of renovating the house, and overestimate the increase in resale value the renovations will bring.
• Secrecy: Once again, because of the secrecy in transactions, the very people who truly know what that value will be – the IT managers who are expected to wring cost savings out of the systems, the sales staff who are expected to bring more profit per sale due to cross-selling, the marketing managers who are expected to drive demand for the better and more complete “whole package” – are left out of the conversation. It is often said that startup CEOs are the best, because they spend real time talking to every single customer, until they can practically step into the customers’ shoes. In larger firms, the sales staff and marketing managers are the ones who spend time living and breathing the customer and must be involved.
• Consultants: Once again, consultants, who have a strong incentive to push a transaction that will bring them huge consulting fees, along with too strong a belief in themselves and their firms’ abilities, push for a transaction by sugar-coating the results. To their credit, it is highly likely that the consultants themselves do not realize what they are doing. It is unlikely to be malicious or intentionally misleading except in rare circumstances. Nonetheless, the damage is done. One of the top brand of consulting firm in the world once gave a strategic presentation, which I was not privileged to attend but did see the slides, in which they showed a “hockey stick” graph of net present value over time. If the two top country partners of this firm do not understand “net present value” and what it means, they cannot possibly understand the true value of a merger.
• Hubris: Every CEO likes to believe that he or she is the next coming, will make their mark as a really huge CEO. In software, Eric Raymond once said (and I paraphrase, errors are mine) that the software is finished not when there is nothing left to add, but when there is nothing left to remove. A really great CEO is not one who knows how to build his/her empire, but one who knows how to focus a business. Like presidents (and I always avoid politics in these posts), the CEO who focuses on his or her legacy does damage to both their charge and their legacy; one who focuses on just doing the best job, is a benefit to their charge and their legacy.
• Incentives: These incentives are the same. If pay is tied to size, rather than profitability, who can possibly blame the CEO or Board for wanting to jumpstart company grow the company 

The only way to get properly estimate these costs and the value gained upfront is to get a fully independent person, working with the board and the line staff (managers on down) to fully figure out the costs, and ensure that the compensation of the CEO is not tied to the overall size of the company, but rather its profitability. Whoever does this analysis – consultant, executive, anyone – must have zero incentive to make it work. They must be paid to figure the value and costs out, and not have millions in bonuses or consulting fees hanging on recommending a transaction. Finally, the staff who live and breathe the day-to-day operations – running the business, marketing its goods and selling to customers – must give input on the true value and costs of the integration.

In an earlier post, we discussed the justifications, at least theoretically, for a merger or acquisition. Of course, like any financial event, whether buying a company or buying an iPod, the numbers have to make sense. Put in other terms, you have to gain more out of it than you put into it (in business terms, being profitable), otherwise you simply do not do it. In this discussion, we will examine some of the financials behind these events, and how they are justified.

In general, you buy something because it is worthwhile to you. If that something is not a simple expenditure (like an iPod), but rather an investment, you do so because you expect to make a return on that investment. The obvious question is, in the case of an acquisition, shouldn’t the net return be neutral? In short, if a company’s profitability (and, more importantly, its free cash), is worth $500MM right now, then shouldn’t its price be exactly $500MM? If so, if you spend $500MM, you have gained nothing; if you spend greater than $500MM, you have lost money, and if you spend less than $500MM, well, the current owners will not sell, since they make more money by holding onto the company, at least until a better buyer comes along.

The core to understanding this is the integration value discussed in the previous article. We will use the horizontal integration example of earlier with two car companies. In that case, the competing companies may well have been worth $500MM each as they stand today, but since each one will cut $5,000, or 25%, of their expenses by working together through economies of scale, when joined each is worth some amount more than $500MM, say, $600MM. Thus, the buying company will pay something more than $500MM, to make it worthwhile for the sellers, and less than $600MM, to make it worthwhile for the buyers. Everyone wins (except for the remaining competitors, who now face a stronger player, but we are not too concerned about them). From a growth perspective, using the cross-selling example, if each company, again, is worth $500MM, but together, without having to cut any costs, they can grow their revenues and hence their profits by 20%, then each one, when joined together, is worth $600MM, or 20% more. Thus, the acquiring company will pay something more than $500MM, to make it worthwhile to the sellers, and something less than $600MM to make it worthwhile to the buyers.

As we can see from the above financial examples, mergers and acquisitions really only make sense in one of two cases:

1. Where the value of integration is great enough that the acquirer can pay more than the current value of the company, yet still have a positive return on the investment, such that the company standing alone could not gain this new value.
2. Where the existing company is so poorly run that the current value is significantly below what the value will be when the new owners turn it around and fix it up.

The final twist in this is the operational cost of the transaction. In addition to the significant sums to pay to acquire (or merge with) the company, there are operational and transactional costs:

1. Investment banker fees
2. Legal fees
3. Regulatory fees
4. Employee severance
5. Facilities mergers
6. Accounting mergers
7. IT systems mergers
8. Benefits management

The list goes on and on.

So why, in the end, do so many mergers and acquisitions fail? The short form, of course, is that the cost paid plus the operational costs is greater than the value the managers/owners perceived they could extract from the merged entity. The obvious question, then, is why are these numbers off for so many transactions? A follow-up post will discuss this issue.

Earlier today, someone posted on a mailing list an attempt to understand why eBay bought Skype in the first place. For those who have forgotten, eBay, the online auction giant, bought Skype, the VOIP/IM newcomer with around 57MM registered users at the time, for $2.6BN in 2005. The poster wanted to understand why they bought Skype, and how they feared Skype upending their business model. This article is a short introduction to the reasoning behind most acquisitions, in an attempt to improve understanding.

First, one very big caveat. Depending on whose analysis you use, over 70% of mergers and acquisitions fail. The definition of failure here is that they do not achieve their targeted goals, and often end up costing more than if they had just let things be. Why do they fail, and given those statistics, why do they try? That has a lot to do with psychology, and will be the subject of a follow-up post.

In general, there are two reasons for entering into an acquisition: vertical integration and horizontal integration. These sound like business-school-buzzwords, which to some extent they are, but they help categorize and explain why a business buys another one.

Vertical Integration

Vertical integration is the process of buying out your supplier or customer, essentially someone in your supply chain. Let us say you make cars (not a very profitable thing nowadays, especially if you are a US car company). You decide, for various reasons, that it will be more profitable for you to own your own windshield maker rather than buy them from one or more suppliers. This acquisition is known as vertical integration. Similarly, if you are Fidelity, and you sell lots of mutual fund via a network of brokers owned by Charles Schwab, and you decide to buy out part of Schwab’s business, that is vertical integration.

Why would a company perform vertical integration? There are several possible reasons:

•  Supply: For whatever reason, you are concerned about supply, and thus you buy a supplier to guarantee good supply. You do not expect cost savings per se; this acquisition is about ensuring a good supply for your business.
• Savings: Each stage of middle-man in a business adds a cost, both cost of operation and profit requirements. If you can perform that function better in-house, then you can realize cost savings.
• Competitive advantage: If there is only one or a few suppliers or customers for a business, buying them can put your competition at a significant disadvantage.

Clearly, vertical integration can be done for cost reasons, but is primarily done for strategic reasons, either to improve your ability to manufacture, deliver and market goods and services, or to make it more challenging for your competitors to do so.

Horizontal Integration

Horizontal integration is the process of buying out another firm that is not in your supply chain, and is either complementary to or competitive to your business. For example, if you are General Motors, and you buy out Ford (difficult to do when you just lost $15.5BN in a quarter), that is horizontal integration. Similarly, if you are Morgan Stanley and you merge with Dean Witter so that each one can feed the other’s business, that is horizontal integration.

Why would a company perform horizontal integration? There are several possible reasons:

• Economies of scale: Quite simply, if you produce 100,000 cars at a cost to you of $20,000 each, and your competitor does likewise, it is possible that manufacturing 200,000 cars together will only cost you $15,000 each, leading to a $5,000 cost savings per car, or $1BN a year. This is one example of that awful buzzword, “synergies.”
• Economies of scope: Similar to economies of scale, it is possible that although you each manufacture, separate, non-competitive items, together you can manufacture each of them more cheaply. This is very similar to economies of scale, except that you are not manufacturing the same item, but nonetheless gaining cost advantage.
• Fixed cost savings: There is a certain amount of overhead in running a firm of a given size. These costs are not lines. For example, in the second quarter of 2008, Google had general and administrative costs (think executives, some facilities, HR, legal, etc.) of $319MM for just under 20,000 employees, yet it would cost much less than double that, or $638MM, for 40,000 employees. Given Google’s notorious inefficiencies in operations, perhaps this is overstating the case, but the principle holds. This is usually what executives of public companies when they refer to “cost savings via synergies” in a vague sense (also usually known as, “we hope we will save some money, else we cannot justify the acquisition”).
• Competitive reduction: You buy a competitor. Using the above example, if GM bought Ford (which is difficult, but then again, Ford lost over $1BN in the second quarter of 2008 and its shares are at around $4.62 for a total capitalization of just over $10BN, so perhaps not so far-fetched), it would have the ability to control a greater share of the market, raise prices or reduce output (if the UAW even let them).
• Cross-selling: You buy a complementary company, expecting that you will each use the advantages of the other to build a bigger business. The whole is greater than the sum of its parts. This was the rationale behind most of the “financial supermarket” mergers and acquisitions of the late 1990s through recent years. Morgan Stanley merged with Dean Witter because they each felt that they could use the other’s strengths and especially markets to sell more than they could on their own. The same holds true for Citi and Travellers, and all the others. It is important to remember that this is not primarily about saving money (reducing expenses), but about increasing revenue (top-line).

Clearly, horizontal integration can be done for either cost or strategic reasons. 

Skype and eBay

So given the above, what type of merger was Skype and eBay? It is difficult to tell. Two primary rationales have been given for the acquisition at various times.

1. Members: eBay saw a slowdown in membership growth, while Skype had tens of millions of active members (around 57MM as of the acquisition). eBay saw Skype as a way to bring many new members into the eBay fold. Essentially, eBay bought Skype for its users. This is straight horizontal acquisition, with cross-selling as a rationale. 
2. Communications: eBay saw a slowdown in membership growth, and believed that Skype, with its popular instant messaging and voice channels, could act as a strong method of communications between buyers and sellers on eBay. Although eBay could have built an IM or VoIP platform on their own, the technical challenges are not insignificant, and the marketing challenges – how many IM and VoIP systems do most users want – are even greater. Essentially, they bought out a good and well-positioned supplier. This is straight vertical acquisition.

Either way, it did not fare well. eBay bought Skype in 2005 for $2.6BN in September 2005. By October 2007, two years later, eBay had to take an impairment charge of $900MM. Put in more normal terms, eBay is saying, “Oops, we seriously overpaid.”

A final note on non-integration acquisitions

As a final note, there is a type of acquisition that has nothing to do with integration. In short, if an investor sees a company that is poorly run, and believes that, without integrating its business with any other business, s/he can run it better, they may buy out the firm to do so. A famous recent example is the 2007 Cerberus Capital acquisition of Chrysler Corp. Chrysler lost $1.5BN on more than $60BN in revenue in 2006 (their last year as a public company). Cerberus took a look and said, “What a mess. We can do better than $1.5BN in losses and growing.” They acquired the company, installed new management (Bob Nardelli, deposed for good reason former CEO of Home Depot), and started a turnaround plan. The Chrysler case will be the subject of a follow-up analysis on this site in the coming weeks. Many of these acquisitions were made famous in the 1980s as hostile takeovers, so-called because the existing Board and management resisted the acquisition directly from shareholders. Considering that this is the same Board and management that mismanaged the company in the first place, and were likely to lose their positions and income as a direct result of the takeover, their rejection was not exactly surprising.

It is a well-known truism that companies lose their market share (and their shirts, and sometimes their life), when they are so tied into their big, currently lucrative business model to support the small opportunities that can grow into big ones. In this instance, I am referring to the credit-card companies who missed the boat on micro-payments, got lucky in that no one stepped in fully (or at least successfully), and seem dead set on doing so again.

Most people think of credit cards in terms of the consumer side. You make a purchase, the agreed price is $x, you swipe your card (or put the number on the Web page or through the phone), and within some period of time get a bill that you pay off. From the seller (a.k.a. merchant) side, it is a little more complex, with all sorts of pricing and fees that you pay. Put in other terms, if I sell you an iPod for $299 and charge your Visa or MasterCard, my store will pay some amount in fees to the credit card company for the privilege of processing the transaction. Why would I do that?

1. It is very hard to do business nowadays without accepting credit cards. This is just a cost of doing business.
2. It is a lot easier and safer to manage my funds via the credit cards, rather than have to worry about cash, which can physically take up space, get ruined, become misplaced, or be stolen, anywhere in my store or on the way to the bank.

The fees paid by merchants vary widely and are very confusing. Many claim this is done intentionally by the processors and credit card companies in order to get more from merchants. If you want a good introduction to how to read the bills, check out the article in the April 2007 issue of Inc magazine. I would also recommend a good credit card consultant if you process any serious volume. In a very simplified version, the merchant essentially pays several fees:

1. Monthly fee: This is a flat monthly fee for the right to accept and process credit cards, and may include a terminal (that card-swipe machine you see at your local store), support or other services.
2. Per-transaction fee: This is a flat fee that is charged per transaction. Average rates tend to be $0.10 to $0.15, according to the aforementioned Inc article and several experts I have spoken with, but tend to move higher for Internet merchant accounts, i.e. those accounts specifically set up for processing payments online.
3. Discount rate: This is anything but a discount. It is the percentage of the transaction the processor takes. It covers both the interchange fees from Visa and MasterCard, as well as profit to the processor. 

 Using as an example PayPal, the most popular processing engine on the Internet, the fees in its two basic plans are as follows. It is important to note that PayPal has dramatically simplified the structures. Most processors do not provide the advanced online services PayPal does, and charge different rates based on card type, etc.

• Website Payments Standard
  1. Monthly fee: $0
  2. Per-transaction fee: $0.30
  3. Discount rate: 1.9% to 2.9%
• Website Payments Pro
  1. Monthly fee: $30.00
  2. Per-transaction fee: $0.30
  3. Discount rate: 1.9% to 2.9%

If you look at the numbers, you quickly see that if your average sale size is $50, $100 or more, the per-transaction fee of $0.10-0.30 is only a tiny amount of the total sale, bumping your discount rate up by at most 1%, and usually far less. No one wants to give up 1% in additional costs, but it is not disastrous. If it is, you have more fundamental business problems and need professional help.

On the other hand, when you make sales of under $10, and especially really small ones like under $1, you can see that the per-transaction fees can double, triple, or worse your discount rate. If you sell 2 readings of an ebook for $5, even if you somehow got a discount rate of 2%, the $0.30 per transaction fee adds 6%, quadrupling your discount rate. This is a serious issue.

These very costs – which used to be worse – are the main reason many stores have those wonderful signs that say, “minimum charge for credit card $10/$20″ or similar.  And it is for this very reason that many businesses built around selling items for small amounts online took a very long time to take off.

How did it get this way? Visa, MC and the like are really just passing their costs on. They have a fixed cost to process each transaction, and they have a percentage cost for each transaction, hence it makes sense to pass it on that way. However, those fixed costs are, largely, minimal. In the early days of the credit card industry, for those of us who remember it, there were no Internet or dial-up terminals. The merchant took your card, made an impression, and then forwarded it on to the processor. Handling all these paper slips was quite a labour-intensive and capital-intensive proposition for both merchant and processor. Add to that the probability of lost slips, meaning lost ability to get the funds as well as possible theft, and the costs could get quite high.

No longer. Today, nearly every credit card transaction is handled electronically. The per-transaction costs are tiny. Visa does not even break it out separately in its 10K. Thus, although these charges may have dropped, they certainly do not reflect the lower costs. Given the efficiencies of electronic transaction processing, the card companies certainly viewed it as in their best interests to reap the rewards of the information technology revolution and earn higher profits.

The issue here is that if they did, indeed, reduce these amounts, the minimum size of a credit card transaction would go down, opening a whole new class of transactions, i.e. micro-payments.

As a result of the credit card firms clinging to their old models, many new attempts to support micro-payments has opened up:

1. Carriers: The wireless carriers have opened up their billing infrastructure. Although their costs are still high, they are certainly more flexible and lower than the credit card companies. Thus, many companies have begun to offer small transactions that would not be economically feasible via credit cards, charging instead via the carriers. Although many of these do eventually get passed through to the credit cards when the carriers charge their bills: (a) any extra middleman means someone is taking a profit they could have had; (b) when AT&T or Verizon wireless, with their massive size, process payments, it is undoubtedly at a far less profitable rate to the processors than the many smaller providers. 
2. ISPs: A number of ISPs began to offer services similar to mobile carriers. This has not taken off in the United States.
3. PayPal: PayPal has been very wise, effectively becoming a bank. Sure, the traditional PayPal is where I have an account with a credit card number, as do you, and I can send money to your PayPal account from mine, thus charging my card. But many people now maintain PayPal balances and send funds from one person to another, or one consumer to a merchant, without ever going near credit cards. The cost to PayPal of these transactions is near-zero, while the lost profit of these transactions to the credit card firms is quite large. Let’s look at it via the numbers. In the last quarter for which information was reported, Visa received $749MM in service fees on payments transaction volume of $652BN. Put in other terms, it makes revenue of 0.11% of processed volume. PayPal, which is both a competitor and a customer of Visa, had $602MM in revenue on $14.93BN in payments volume. In similar terms, PayPal makes revenue of 4.0% on processed volume, or 36 times the revenue per dollar of transaction. Some of this is simply due to PayPal being in a higher-fee business, further down the food chain from Visa. But a large amount of it is due to PayPal being able to simplify transactions between accounts without needing the credit card companies.

Other similar firms are slowly (or not so slowly) popping up.

The credit card companies missed the boat. They viewed themselves as indispensible, that no one can do non-cash or slow-check transactions without them, thus:

1. We can charge what we want.
2. We can structure it how we want.

Along came the carriers, and especially the PayPals, and said, “we will change the model”:

1. We will provide low-cost person-to-person or business-to-business transactions without the credit cards
2. We will make it a “push” model: the money is transferred when I push out to my merchant, rather than giving the merchant enough sort-of-secret information, like my credit card number, known in industry parlance as PAN, and lately lots more information, like the address, secret number/CVV/CSC/CVV2/etc., wherein they now can “pull” information out of my account. I recently was involved in a PCI compliance project with a company that processes large amounts of credit card transactions each month.

What will happen next?

1. Smaller competitors (assuming PayPal and similar to be “small”) will continue to provide more secure, easier-to-use, easier-to-protect and lower-cost models, eating into the smaller, and even larger, transactions that used to be the sole domain of the credit card industry.
2. These competitors will begin to see how they can provide better solutions even for larger transactions and will begin to offer methods to make payments not only online, but directly compete with the cards’ bread-and-butter, “card present transactions.”
3. Eventually, some of these may dwarf or even acquire a credit card company. Certainly the turmoil in the financial services industry, between the credit crunch and corporate troubles, will leave at least one weak enough to be acquired.

It is hard for those of us who remember the radical notion of a credit card, “will that be cash… or Chargex,” to think of the credit card industry as a dinosaur. But the reality is that it is an old-time business. Unless it changes its model, it will be in trouble soon.