Symphony and the Big-Firm Technology Credibility Problem

In its quest to win over financial professionals, new communications platform Symphony faces a familiar challenge.


Symphony, the new messaging network financed by a consortium of 15 buy- and sell-side institutions, has worked strenuously since its launch late last year to dispel the widely held perception that it’s a hotbed of conflicts of interest. The project, as I described in a story last week, is being led by former Skype and Microsoft doyen David Gurle, a likable Frenchman with a quick wit and a deep love of commercial communications networks. Symphony has always had to contend with its difficult origins: It started as an in-house technology project at Goldman Sachs Group in 2013, the year that the relationship between the bank and Bloomberg, whose fabled terminal is the nerve center and communications hub for most of what happens in the elite front office of the global financial industry, broke down amid the improper use of terminal data by Bloomberg News reporters to generate stories about Goldman employees. Symphony and Goldman stress that work on the new communications project had already begun before May 2013, when news of the spying scandal broke. But in the minds of most financial professionals, Goldman started work on what would eventually become Symphony in direct retaliation for the incident.

To combat the perception that Symphony is a Goldman vehicle, the bank did two things. It enlisted 14 other financial institutions, including Bank of America and BlackRock, as members of an investment consortium to fund a new vehicle to handle the development and commercialization of the new messaging network, then contributed to a total consortium-wide venture-capital-style investment, in late 2014, of $66 million into that new vehicle, Symphony Communication Services. (Symphony won’t say how much each of the 15 consortium members has contributed to that $66 million.) By spreading the equity across a broad group of investors and making clear that Symphony was a business built to scale and complete follow-on financing rounds in the manner of a regular tech start-up, Goldman hoped to show this wasn’t purely a “Goldie project.”

Gurle himself has stuck hard to this script in the six months he’s been the CEO of Symphony. When I spoke to him earlier this month, he was keen to emphasize three points. First, Symphony is not a business designed to build communications tools for Wall Street only; once he’s done completing his ambitious plans for messaging in finance, he hopes to expand the offering to other industries with similar communications problems, such as retail and health care. The business’s location — its head office is in Palo Alto — emphasizes how Symphony has situated itself in the mainstream of the heavily populated software-as-a-service tech start-up world: This is a company that aims to appeal to a broad audience in the business world, not just sit in the shadows of Wall Street and cater exclusively to the trading desks and pitchbook factories of the financial industry.

The second point is that Symphony is not designed to take Bloomberg down — not in a frontal sense, at least. Gurle is well aware of the long history and stellar record of failure of all the “Bloomberg killers” that have emerged in the 30-plus years since the former New York mayor launched his namesake financial data empire. But he concedes that one of the reasons Symphony, which has 50,000 users today and hopes to reach 100,000 by the end of this year, was conceived and financed was because many large financial institutions don’t trust Bloomberg. And it’s also clear that once his grand vision for Symphony is realized, the product that eventually emerges will include features in sufficient number — Gurle name-checks content, workflows and an App Store for open development of new financial analysis tools by external developers — to begin to make it a match for the enormously information-rich Bloomberg terminal, which includes more than 15,000 functions.

Gurle’s third point illustrates the degree to which Bloomberg, for all of the claims that Symphony’s offering is at once more specific and more ambitious than the terminal, is on his mind. The Symphony CEO repeatedly emphasizes the protections that his firm, which rolled out its beta version last week and hopes to have a general audience release in mid-July, has put in place to ensure the privacy and the confidentiality of communications on its platform. Privacy is a key concern for financial professionals, since a lot of information relating to order flow, including transaction and price data, gets transmitted through communications networks (chiefly, today, the Bloomberg terminal’s IB chat tool). Gurle says Symphony has built its network with “end-to-end encryption” over which end users have complete control. This makes the system secure as a matter of “technology, not policy,” he claims — a none too subtle dig at Bloomberg, which revamped its privacy procedures in the wake of the 2013 spying incident, following a review led by former IBM CEO Sam Palmisano. Gurle’s point is that even if some fictitious bad actor within Symphony or its investor consortium wanted to compromise the privacy of users’ communications, they would not be able to: Technology, not voluntary policy, would prevent them from doing so.

A diverse set of investor interests, the desire to scale beyond finance and watertight privacy protection technology: With all that in place, surely you’d think the financial industry could set aside its reticence and give this interesting new initiative the benefit of the doubt? Responses and messages I’ve received in the week since I published my story suggest things won’t be that easy for Symphony. For a start, many financial professionals question whether having 15 investors in the initial consortium will dilute the focus of operations: “That’s a lot of cooks to keep happy,” one trader, who requested anonymity, told me. There’s also the simple issue of why a business built to scale and create products in industries beyond finance has started off with a consortium of exclusively finance industry investors. This would be a highly unusual move for a broad enterprise-focused start-up seeking finance through the regular venture capital channels.


But the real and biggest problem is that traders, for all Gurle’s sensible and seemingly persuasive assertions, don’t trust Symphony. This distrust is not born of a deep sense of loyalty and affection for Bloomberg; most financial professionals see the market data industry as a monopoly (or, at the very least, once Thomson Reuters is added to the picture, a duopoly) and would welcome stronger competition. But for now, many traders appear hesitant to embrace Symphony for one simple reason: They continue to see it as a Goldman vehicle.

Late last week one trader at a midsize hedge fund, also requesting anonymity, sent me a long and richly detailed e-mail with his thoughts on the whole issue. “There are many things on Wall Street to lol about,” he began, adopting the Internet lingo of a person possibly one-third his age, “but Goldman thinking that we would trust them with our private messages instead of Bloomberg is absurd. Everybody on my side of the Street is afraid of trading with bulge — and specifically Goldman — because they know their orders are being used.”

In a previous job, the trader worked for a large buy-side institution. He continued:

“We ran over 3 billion in illiquid small caps. We traded a lot through Goldman, and they were by far most by our most efficient broker. This was much to our amazement, since their focus is (we thought) on large-cap stocks. So we wanted to do more business with them, but first we had to examine them more closely: Like everyone else, we assumed they were trading against us.

“So I was sent to New York to spend a day on their small-cap desk. I’ve been on the Street for 20 years, and I’m quite intimate with the inner workings of the broker-dealer complex, so I was an informed observer. They wanted to impress me with their Chinese walls, and even though they may have executed our trades quite efficiently, if for example Fidelity came in to sell 30 million shares of Cisco, Goldman equities would get it done quite well. But you bet your ass that somewhere else at Goldman, on another floor in the same building probably, somebody immediately bought puts on Cisco.”

The trader concluded, “It’s hard enough sending them our orders, knowing they’re milking us in some way. God knows what they would do with our messages.”

To be clear, I’m not implying that anything improper happened in this instance. Nor am I implying that Goldman misuses client information or order flow. But it’s a question of perception, and that is the nub of the problem — not just for Goldman with Symphony but for any large financial institution that takes positions in the markets and seeks, simultaneously, to offer its nominal market competitors technology services as a separate business. If a dominant, market-making firm like Goldman gains insight into the trading intentions or deal flow of one of its putative technology customers, that information can be very valuable to Goldman — and very detrimental to the customer. Goldman can use it to trade against its own client. By contrast, the worst that can happen if Bloomberg, which does not make markets, gains access to that information is that a reporter can go off and write a story: a potential source of ephemeral reputational damage to the customer, certainly, but hardly the basis for a big trading loss.

Several big institutions with deep trading books have struggled with the basic problems of trust and credibility that arise whenever they want to sell technology services to trading counterparties. The creators of REDI, a brokerage and trading technology platform that Goldman launched on its own in the 2000s, and Citadel Execution Services, the automated market-making platform started by Chicago hedge fund firm Citadel in 2005, both faced major hurdles in their early years to convince the Street of their basic trustworthiness. Goldman eventually sold a majority stake in REDI to a group of Wall Street banks, whereas Citadel Execution Services has gone on to cement its status as one of the U.S. equity market’s principal execution venues. It’s clear that Gurle is thinking hard about how to respond to the broader financial industry’s reluctance to entrust communications to Symphony. But it’s equally clear there’s much work still to be done.