Download PDF

There is a common narrative thread in the fixed income markets today that essentially says, “the fixed income market today is like the equity market of 20 years ago (or 10, or 30).” The implied conclusion is that in 20 years (or 10, or 30) the fixed income market will look just like the equity market of today.

While in most ways this a stretched analogy, in one meaningful way the narrative holds true. In 1998 when I joined the global equity portfolio trading desk at Merrill Lynch it was pounded into my head that, “It’s about moving, not storage.” Wall Street loves a catchy phrase, and for good reason. I am still quoting it 20 years later. We were changing the mindset of the traders and sales people. Earning trades through capital commitment (“buying the trade” aka storage) was no longer acceptable. It was our jobs to earn agency business for the firm (moving). Collecting commissions rather than taking risk was the future. (I only wish that mantra had existed in the mortgage securitization department a decade later.)

The feedback I have heard from a few head fixed income traders lately is that they no longer get paid to execute with the street (i.e. manage inventory). Instead their priority is to fill customer orders. Moving, not storage. The idea that a trader might identify a “cheap” bond, take it into inventory, then try to find a home for it with a customer, but failing that, be able to sell it back to the street at “fair value” on some future date is becoming an antiquated notion. Moving, not storage. Perhaps this shift in business model was started by the post 2008 financial crisis regulatory changes that make capital very dear for brokers. It is ironic that this is the case in an age where capital is so cheap for everyone else. One theory goes that we have a record number of unicorns in the United States because it is so easy to raise money in private markets without a founder losing control that it no longer makes sense to do an IPO. And even some IPO structures are reflecting that reality with management selling stock with limited or no voting rights attached.

If the moving, not storage business model is truly here to stay in fixed income markets (it certainly appears to have become more ingrained in equity markets since 1998) then, following the equity market example, the next logical steps are:

  1. A division of labor between those who commit capital and those who face customers.
  2. The capital committers will “Wal-Mart” the market by accepting tighter margins to gain bigger market share.
  3. The customer facing firms will focus on a “just-in-time inventory” approach.

Division of Labor

Customer facing firms will lean on the capital committing firms to make the markets while they service the customers. After all, customer acquisition is a very expensive proposition. The firms that already have the customers are most likely to keep the customers. Recently, and to great fanfare, robo-advisors attempted to dislodge existing customers and grab new customers that were just entering the industry. What they found is that the cost of acquiring the new customers was often greater than the expected lifetime value of the relationship. As for dislodging existing customers Vanguard, Schwab and Fidelity have very successfully prevented that from occurring by offering their own versions of the same products. And of course, they can do this cheaply because it is less expensive to keep a customer than it is to acquire a customer.

Wal-Mart (or Amazon, take your pick)

Capital commitment will become dominated not by the ability to find value, but by the ability to manage risk. Managing risk is a mathematical exercise and a game of scale. The drive to gain scale leads to a tightening of bid/ask spreads. Tighter bid/ask spreads drive smaller competitors from the market place. Call it the Wal-Mart effect or the Amazon effect. If the market is divided among many small competitors, they each need nickels per trade to survive. Fewer large competitors require only pennies. There are definite economies of scale involved in managing trading inventories, if only by reducing industry wide overhead. Something similar has already happened in equites and especially in ETFs. In the municipal bond market there are brokerage firms that have no retail customers. Instead they regularly bid on 10,000 small trades per day for other brokers who do have retail customers. They are pure liquidity providers whose only complaint is that they do not win enough of the trades they bid on. Their collective drive for scale will tighten bid/ask spreads for bonds. The danger is that, like in equities, they will ultimately only want to focus in the most liquid end of the market.

Just-In-Time Inventory

With one source of revenue gone (buying “cheap” bonds from the street) firms need to maximize the other source of revenue, servicing the customer. To do this will require greater levels of understanding of their customer base than ever before. But this understanding cannot be housed in the individual brains of the individual advisors. To truly maximize the value of these relationships requires a firm level understanding of the individual customers. Given the size of these businesses this can only be achieved through greater automation. Firms cannot afford to let a single opportunity to service the customer slip through their fingers. If they are not automating alerts of cash building up in portfolios, bonds maturing, or being re-rated, individual bonds no longer fitting the desired portfolio they are a part of, or a hundred other factors that affect customer satisfaction then they are not optimizing customer outcomes. Brokerage firms have invested heavily in building an ecosystem for their customers. Uncovering the latent demand in that ecosystem efficiently is the key to their continued success. Those with the best software win.


Coming back to our “bonds are like stocks 20 years ago” analogy, it is unlikely that we will have a Reg NMS environment for bonds in 20 years. Retail customers do not participate in fixed income price discovery the same way that equity investors do. Essentially, they do not typically view bonds as trading instruments. When retail customers talk about “playing the market” they are always referring to the stock market, not the bond market. Retail customers look to the bond market for income and safety. And service. The way the bond market of 2038 will resemble the stock market of 2018 is the division of labor between those who move and those who store. And the successful movers will have the best tools for servicing their customers.

Keep on reading: