Soft dollars–what they are and how their demise will affect professional equity managers

What they are

Soft dollars, sometimes also called research commissions, are commission or bid-asked spread payments (for over the counter securities) made at a higher than normal rate by investment managers to brokers as compensation for research services the broker provides.

Although investment managers are required to obtain the lowest cost and the best execution in their trading for their clients, the SEC created a safe harbor years ago that allows soft dollar commissions to be paid, providing that:

–services are genuinely for research, and

–clients understand this is going on.

The SEC has publicized the soft dollar issue a number of times and has given illustrations of types of services that are (like the broker’s own research, Bloomberg terminals, third-party research) that are allowable, as well as services (like transportation costs, furniture, rent) that are not.

Why is this an issue?

Two related reasons:

in paying commissions, the manager is using the client’s money, so he has an obligation to spend it wisely; also,

by using soft dollars, the manager is using client money for expenses that would otherwise come out of his management fee.

Who understands this practice?

Institutional investors understand the use of soft dollars very well.  In fact, some require their managers to direct a certain amount of commission business to brokers that the client designates, to satisfy soft dollar arrangements the client has entered into.  Retail investors?–not so much, I think.  For example, do you know what policies your mutual funds follow?  or how much of your investment money goes to buy services your manager uses?

Sizing soft dollars: a percentage of commission payments

Consider an investment management company that has $10 billion in equity assets under management.  If the stocks are all US (commissions are typically higher abroad) and the average price of a share of stock in the fund group’s portfolios is $40, then the group holds 250 million shares of stock.

Let’s say that the group turns over 80% of the portfolio yearly and that it pays the following commissions (including imputed commissions for NASDAQ trades):

electronic crossing networks        20% of volume                   $.015/share

soft dollar trades                               20% of volume                   $.08/share

“regular” brokerage trades             60% of volume                   $.04/share.

If we figure this out in dollars, the amounts are:

crossing networks               $600,000         11% of total commission expense

“regular” trades                 $1,600,000         30%

soft dollar trades               $3,200,000      59%.

A soft commission dollar doesn’t buy a hard dollar’s worth of goods or services.  A typical ratio is that 1.7 soft dollars equals one hard dollar.  Using this relationship, our firm’s $3.2 million soft dollars buys $1.9 million hard dollars worth of research services.. a boost to the manager’s profits

Let’s relate this to our hypothetical firm’s profits.  If we assume that the firm charges an average fee of 50 basis points on its $10 billion in assets and has a relatively constant 50% operating margin (I’m not saying this is the best way to analyze any company, but it is fast…and gives a reasonable ballpark number), then the company has operating income of $25 million.  Using soft dollar power supplied by client commissions raises the company’s operating earnings by about 9%.

The CEO of our investment company might comment that this $2 million adds up to only .002% of the assets under management, so that this practice really doesn’t affect investment performance by any noticeable amount.  (This would be sort of like saying that taking money is ok if you don’t get caught–not exactly a confidence builder in the speaker’s attitude toward fiduciary responsibility.)

On the surface at least, soft dollar trades appear to be no more or less profitable than “regular’ trades.  If processing costs are about $.015/trade, a $.04 trade yields a profit of $.025.  An $.08 soft dollar trade has $.047 in costs to provide services (figured at the retail price the investment manager would otherwise pay) + $.015 in processing costs,  so it yields a profit of $.018.  But brokers are in a position to bargain with suppliers for quantity discounts.  The SEC investigated the discount question about ten years ago and found that most brokers supplied the services at their cost, but that some charged a distribution fee.  Bernie Madoff has subsequently unveiled the limits of the SEC’s investigative skills, so it’s a safe bet that a short conversation with service providers would reveal that brokers universally take a substantial markup.

Why managers don’t highlight this practice

I’ve written enough, I think, to show why investment managers aren’t falling all over themselves to explain their use of soft dollars to their customers.

Their marketing material is typically filled with references to their wide-ranging research efforts, their unique insights and their proprietary processes.  However true they believe this is, nevertheless a substantial portion of what they do involves purchasing research information from brokers and third parties–information that any of their their rivals can (and do) buy as well–spending millions of dollars of client money yearly doing so.

But that’s not the main problem with soft dollars, in my opinion.

When I entered investment management (over 30 years ago), the business model was to have in-house research, supplemented by input from brokerage analysts.  Buy side firms either trained their own securities analysts or recruited them from the the ranks of brokerage house industry analysts who had become burned out from the constant marketing travel and presentations/dinners/conferences that sell-side positions require.  There were very few third party research efforts.

As the business grew, and as it matured (i.e., the market began to be saturated and growth slowed), management firms realized that they could expand profits at a faster rate if they used brokerage research, purchased with soft dollars, not as an adjunct to their own research efforts, but as a substitute for them.

The old model has morphed today into one where many investment management analysts do little original work of their own but collect, synthesize and comment on the output of sell-side analysts instead.  Layoffs of brokerage house analysts in the wake of the bursting of the internet bubble and of the financial crisis have spawned a whole raft of third-party research services produced by analysts well-known to the buy side.  Investment management companies purchase these services, too, through soft dollars.

Although this isn’t the way I would run an investment company, an outsourcing strategy makes some sense.  It’s hard to generate consistently good investment performance with today’s super-large portfolios.  It’s probably easier to produce strong inflows of money with snazzy marketing.  Also, as an industry matures, the CEO tends less to be a technical expert at manufacturing the company products, and increasingly a savvy marketer of an increasingly commodity-like offering.  Unfortunately for its users, though,the outsourcing model seems to be coming apart at the seams.

Brokers no longer want it

For at least the past twenty years, the brokerage industry has regarded research as a loss leader, needed to get investment banking and money management clients in the door, but an area to be the object of cost control rather than expansion.  Two cyclical downturns in the current decade haven’t enhanced the outlook for brokerage house researchers.

In addition, the stock trading business is well along in a secular switch away from brokers to electronic crossing networks, which may not have as much liquidity but which offer greater confidentiality for the user and very much lower costs.  That is, the market is shrinking.

Three consequences:

1. lots of experienced analysts have left the brokerage houses for hedge funds, which continue to value proprietary research;

2. in response to lower overall profits, brokers have laid off a substantial number of their remaining research staffs, some of whom have formed third-party research boutiques;

3. brokers are increasingly rationing their remaining research output, based on who pays them the most.  Hedge funds, which are not subject to the same SEC regulation on payments to brokers as most other money managers, are high on their lists.   As usual, it seems to me, the retail client and/or his financial advisor are the caboose in the line to get information–and sometimes get nothing at all.  In today’s world, however, all but the top few money management firms are deep in the back end of the queue as well.  (More on this in my next post.)

Fidelity also wants to end soft dollars

It’s also important to note that Fidelity, which I think single-handedly sounded the death knell for the load mutual fund business during the Eighties by severely reducing or eliminating sales charges on its funds, intends to play a leading role in stamping out the soft dollar business in this decade.  Why?  Ending the practice will likely do severe damage to the reputation and profits of rivals (meaning almost everyone else) who have come to depend on it.

In a game-changing move that has generally been forgotten during the financial crisis, Fidelity began to negotiate with its brokers to obtain full access to their research in return for a cash payment made from Fidelity’s own money, not the client’s.  It reportedly reached agreements with Lehman and Deutsche Bank for a yearly fee of about $7 million each.


1.  Many other investment management companies benchmark themselves in their soft dollar usage by Fidelity’s practices, which have become the de facto industry standard.  But if Fidelity’s soft dollar commissions are 0%, how can someone else have anything other than 0%, too, especially if he has spent the past fifteen years justifying his soft dollar use on the rule “do what Fidelity does.”

2. Very few other companies could afford to pay, say, ten brokers $7 million each for research services.  In the example, we’ve been using, the firm cuts its profits in half–and its stock price if it’s public–if it pays two brokers this way.

3.  Fidelity has a boatload of sector funds, where budding analysts and portfolio managers can, and have, cut their teeth.  So ending soft dollars doesn’t bother them.

4.  If my description above is close to the mark, and I think it is, many smaller firms have half forgotten how to run comprehensive equity research in-house.  Recapturing that skill–changing the corporate culture–won’t be easy.  Neither will deciding to spend an extra $2 million on in-house research (our hypothetical firm) during an industry downturn to replace soft dollars.  It would also be interesting to discover how much a firm’s board of directors really understands about its soft dollar expenditures.

4.  Brokers may have seen the handwriting Fidelity has put on the wall.  That would be a reasonable interpretation for the surprisingly high (to me) level of layoffs of high-priced securities analysts by the brokers during the current downturn.  So there may not be that much valuable research to buy with soft dollars–and the effort to find it in research boutiques may be very time-consuming.

Leave a Reply

%d bloggers like this: