Is Goldman a dinosaur?November 2, 2010: 3:00 AM ET
Goldman Sachs claims it can still deliver big returns. But new post-crisis rules mean it can't operate the way it used to. The surprise: There may be hope for investors.
Over the past two years Goldman Sachs's once-exalted public image has been battered, its conduct pilloried by politicians and the press. In July the investment bank agreed to pay a humbling $550 million to settle Securities and Exchange Commission charges that it misled investors in a collateralized debt obligation (CDO).
But for all the controversy surrounding the bank, there has been one point on which even its fiercest critics and staunchest defenders can agree: Goldman is a money machine the likes of which has rarely been seen. Last year it earned an astonishing $12.2 billion, vs. $8.8 billion for the dramatically bigger J.P. Morgan Chase (JPM) and losses for Citigroup (C) and Bank of America (BAC). And Goldman (GS) has been promising investors that its profit train will keep chugging at top speed in the years ahead -- a 20% average return on equity (ROE, earnings divided by the difference between assets and liabilities).
The problem is that Goldman's greatest advantage and signature specialty -- its prowess at trading -- is being drastically curtailed. The financial crisis brought a web of new regulations, many of which will hamper Goldman. Proprietary trading will be largely off-limits for the company. (Indeed, in late October news broke that a group of Goldman prop traders is decamping for Kohlberg Kravis Roberts.) Meanwhile, the new rules are limiting the use of debt and forcing a lot of lucrative derivatives trading onto public exchanges. All of that, as we'll see, will shrink profits. "Reregulation means lower risk and lower returns," says Moody's analyst Peter Nerby. "Goldman hasn't reset its promises to shareholders. In this environment, Goldman will be hard-pressed to deliver a 20% ROE."
Goldman executives routinely describe themselves as "the fastest animals on the savannah" in conversations with analysts. Braggadocio aside, Goldman deserves at least some of its mystique: It has plenty of capital, and its sober, conservative approach to risk management steered it through the credit crisis in relatively strong shape. The bank is profitable and will remain that way. But when it comes to generating outsize growth in the new world, it appears, the bank's business model may bear more resemblance to a dinosaur than a cheetah.
The slowdown is visible already. Earnings for Goldman's third quarter sagged 43% compared with the same period in 2009, and ROE came in at 11.2%. Certainly, it would've been hard to match 2009's record performance. But one fact is inescapable: Goldman will find it nearly impossible to meet its projections, which will require that it average 24.5% annual earnings growth for the next several years.
Yet even if Goldman can't deliver, here's the paradox: The stock market appears to be overreacting. Investors are now treating its stock as if the bank were a wheezing smokestack company with anemic growth prospects. Its share price is so low that it's actually a pretty good investment.
Goldman spokesman David Wells declined to comment on Fortune's analysis. But in recent earnings calls and conversations with analysts, Goldman executives have stuck with the bank's projections. They argue that it enjoys excellent growth prospects and point to Goldman's leading market shares in both trading and investment banking in Asia, one of the great financial markets of the future. They also claim that much of Goldman's core market-making business will become even more profitable: The new capital requirements will drive competing banks to shun risk, leaving Goldman with a bigger share of the most lucrative trading business and higher margins on those difficult-to-execute trades.
No matter how you analyze it, trading has become crucial for Goldman. In 2000, trading -- both with its own capital and as a market maker for customers as varied as hedge funds and insurers -- accounted for about 50% of Goldman's revenues. Today it's running at 75%. Goldman is the last major bank to depend primarily on this volatile business.
By far the biggest part of that franchise is market making for clients. Goldman buys stocks, junk bonds, other securities, and derivatives from insurers, pension plans, or mutual funds, then holds them while it finds a buyer. Goldman is essentially a middleman: It sets a target price called a "bid" that it attempts to buy at and then presents buyers with a proposed selling price, or an "ask." The wider the spread between bid and ask, the higher the profits for Goldman.
From 2002 to 2007, customer trading proved a fabulous growth business as equities soared, interest rates fell, and global economies grew briskly. Goldman became the biggest player by being more daring than its rivals. It won new customers because it was willing to buy huge blocks of their securities, including exotic instruments that were difficult to sell. At the same time, Goldman skirted losses through adroit risk management.
Goldman's trading took off also because the firm used borrowed money to increase the securities it traded (its so-called trading book). From 2000 to 2007, the ratio of its market-making portfolio to equity grew from 14 to almost 19, and Goldman expanded its trading book fourfold, from $167 billion to $646 billion.
The increase in leverage means Goldman's trading inventory rose more swiftly than its equity. As a result, it made rising returns on equity even as its trading results stayed flat -- typically, 3.5% to 4.5%. From 2002 to 2007, its ROEs averaged an astounding 27%, driving an explosion in earnings.
Then came the crisis and tough new rules. The Dodd-Frank banking legislation and the Basel III rules, which set capital requirements for the world's major financial institutions, allow far less leverage than Wall Street firms used in the past. Just as adding leverage once increased profits, reducing it will lower profits. To reach its targets, Goldman needs to generate enormous revenues on a far smaller trading book than in the past.
That's clear if you're willing to engage in a little math. Goldman is likely to average about $68 billion in common shareholder equity next year. So it would need more than $13 billion in after-tax earnings, plus another $640 million to cover its preferred stock dividend, or about $14 billion. That translates into $21 billion or so in pretax earnings. As we'll see, it's likely to generate about 60% of that.
Those earnings need to come chiefly from Goldman's dominant trading franchise. But let's start with the bank's other two businesses: asset management and advising on mergers and acquisitions. At their 2007 peak, those two businesses generated total revenue of about $15 billion. The company consistently ranks No. 1 in M&A advisory revenues, and rates in the top three in equity underwriting. But those franchises don't grow rapidly and are too small to produce anything resembling the numbers Goldman requires. Let's project that as global economies rebound, M&A and asset management return to their peak revenues of $15 billion. At typical margins of 30% or so, those businesses would generate pretax profits of $5 billion at most.
So trading needs to provide $16 billion in profits to reach Goldman's target -- and that seems nearly impossible. It means that the bank would need $36 billion in revenue from that franchise, since it generally delivers margins near 45%.
In its glory years Goldman was making around 10% of its trading revenue -- between $2.3 billion and $3.8 billion from 2005 to 2007 -- from proprietary trading. Those gains came in part from taking big positions alongside its investors in private equity funds that it raised. But the Dodd-Frank bill puts severe limits on future proprietary trading. Goldman is gradually reducing its principal investments. For now, they should produce around $2 billion a year.
That means Goldman needs to extract $34 billion in revenues from market making. If Goldman's leverage were expanding, that might be achievable. But its leverage has dropped from 19 to eight, and its trading book has fallen by 30% to around $450 billion. Goldman would need a return on those assets of 8% to achieve the necessary trading revenue.
That's almost twice the 4.5% that Goldman achieved from 2001 through mid-2010. Remarkably, the bank actually hit that 8% number, or close to it, in 2009 and early 2010.
But that was the product of an extraordinary market for Goldman, one not likely to reappear for a while. In 2009 the credit crisis was fading and big investors were suddenly able to shed securities that buyers wouldn't touch in 2008. The investors were so anxious to sell, in a generally rising market, that they didn't mind that Goldman bought their securities for a nice discount, then steeply marked them up. Spreads and volumes swelled. Meanwhile, prices for stocks, bonds, and most commodities rose, so Goldman made out there too.
What's a realistic projection for Goldman? If the company generates its usual 4.5% return on its trading book, its market-making revenue would total $20 billion or $21 billion. The return on equity would be about 12%, and 2011 after-tax earnings would be in the vicinity of $8 billion. That's roughly what the bank is slated to earn this year, according to most analysts. It's one-third less than Goldman made in 2009, but a more than respectable level of profits.
The good news, according to valuation expert Steve O'Byrne of Shareholder Value Advisors, is that if Goldman delivers those sorts of profits, investors will reap a total return of 14% per year for seven years and 10% thereafter. That's not shabby, although even that performance is far from guaranteed. Analyst Mike Mayo of Crédit Agricole Securities has a slightly more optimistic outlook: "Goldman will not achieve anything like a 20% ROE, but it can achieve better ROEs than the market is pricing in. Hence, the stock is undervalued." Wall Street's perpetual overachiever has become -- as a stock, anyway -- an underachiever.