Bankers Beginning to Look a Smidge Desperate: Goldman Looking for Technology Magic Bullets to Fix its Cost Problem

This is Naked Capitalism fundraising week. 468 donors have already invested in our efforts to shed light on the dark and seamy corners of finance. Join us and participate via our Tip Jar or WePay in the right column or read about why we’re doing this fundraiser and other ways to donate, such as by check, on our kickoff post or one discussing our current target.

Now that Wall Street blew up the global economy in its search for fun and profit, it is finally having to eat its own cooking in the form of more modest profits. Of course, the slightly chastened Masters of the Universe seem constitutionally unable to recognize that their own actions might have something to do with the fix they are in. Yes, the immediate aftermath of the crisis looked just dandy, as super low interest rates and official confidence building led to some lovely trading opportunities, which in turn produced record bonuses in 2009 and 2010. But as the Fed flattened the yield curve and the economy has stayed stuck in low gear, investors aren’t keen on taking risk, even if negative real interest rates leave doing nothing as an unattractive option.

An article at Dealbook gives some insight into the distress on Wall Street (yes, Virginia, uncertain bonus prospects have this way of focusing the mind in banker-land). Lloyd Blankfein predictably makes “new regulations” the top scapegoat for Goldman’s reduced profit outlook, along with an uncertain market outlook. Ahem, the securities business has always been highly cyclical, but the Greenspan and Bernanke puts did such a good job of dampening the downside that the industry seems to have forgotten what a normal downturn looks like, let alone the sort of barely-averted Great Depression that we are working through.

So the current party line is that Goldman will be using more technology and cutting expenses to cope with the new normal. That sounds great until you dig into the nature of IT at big trading firms. There is a long, seldom told history of ambitious projects gone awry, as in multiple hundreds of millions of dollars poured down sinkholes. I’m hardly close to it, but even I know of a a major effort to build a fixed income utility in the 1990s was a total loss. Goldman had another major fixed income initiative in the later 1990s, to build a fixed income client portal. My understanding is that that was a costly flop. And remember, Goldman really is pretty well managed by securities industry standards.

Blankfein did make a sensible comment that was picked up in the Financial Times writeup of the same presentation: “Our industry has a long history of letting too many people go at the bottom of the cycle and hiring too many at the top.” But I heard firms making the same sort of statements in the last really bad downturn (the early 1990s) and they all (even Goldman) eventually started getting rid of people that they would have liked to have kept (this was acknowledged by headhunters: being fired in that era was not stigmatized if you were mid level. The firms were keeping the senior folks and the cheapest, the analysts and associates, and going through everyone in between with a howitzer).\

But how does IT help ameliorate this situation? Blankfein wasn’t terribly specific except in given the impression that the firm intends to use more automation on the sales and trading side of the business. From the New York Times:

In the firm’s equities business, for instance, Mr. Blankfein said that 60 to 70 percent of shares are now traded through “low-touch” channels, without the direct involvement of people. Electronic execution accounts for significant portions of activity in the firm’s cash fixed-income business as well, Mr. Blankfein said at the conference, which is intended to give Wall Street analysts a broad look at the industry.

The commitment to new technology is part of a broader effort to streamline the investment firm. With regulators imposing stricter requirements on big firms, the former ways of Wall Street now look less desirable.

“For the first time, it’s clear that size and complexity come with a higher cost,” Mr. Blankfein said.

An ex Goldman informant gives a sanity check:

I only know bits of the history but, to me too, this does look like standard-issue CEO BS about IT.

In fact GS’s 70% zero touch, or anyway “low” touch, doesn’t sound that spectacular. I sorta think I remember that, in equities, GS had already attained that level when I was there; a decade ago now.

But perhaps Blankfein’s definition of “low” touch is more stringent now, or perhaps he excludes touches by Ops. Or he isn’t including GS’s voluminous Prop Trading side (or whatever they call it now, presumably not that any more), which has been heavily automated for a long time.

Most likely the Ops side just got left behind in the budget stakes by the sales and trading binge of the noughties, and now looks more than usually precarious and expensive. That’s what usually happens in the boom and aftermath. Time for a huge failed Ops IT project again, I imagine.

Back in the 01/02 downturn the IT horror was GS’s very ambitious all-in-one web-based Prime Brokerage offering, which, looking back now, must simply have had the effect of exposing, to a pack of talkative and demanding hedgies, glitches in some of GS’s really not so bad, but still not 100% reliable, internal systems. Cue high level management concern…heads rolled…

In the infrastructure downstream of trade execution I doubt if all that much has changed in the last ten years, and that is what will keep the level of ‘touch’ obstinately high. Lloyd probably still thinks it’s all about execution, not downstream processing. That’s what I would expect: he’s a trader.

However the amount of investment that would be needed to make a serious dent in the remaining high touch 30% is pretty daunting, and it’s as much to do with middle and back office as with execution. Getting to low touch is obviously all about making the trade lifecyle data driven.There’s no mystery about that, but actually doing it gets terribly hard, quite early on. For instance to make progress from the 70% figure one would probably have to rework, more or less radically, international settlement and custody interfaces, stock master data, and counterparty master data. One would also modify or rebuild the numerous front, middle and back office systems that actually use the data. For a big bank, that comes to many $100millions and north, a size of project that always screws up; far too many moving parts, too much risk with core IT, and all for a cloudy and remote ROI anyway. That is part of why banks gravitate towards tactical solutions, not strategic ones.

In fact often you’ll find that the data that you need in order to automate simply doesn’t exist. Creating the data typically requires other banks, and other players (custodians, exchanges, even legislators, for instance) to move in a similar direction at the same time. I bet it never happens.

Straight through processing has been an industry buzzword for as long as I can remember. I think it’s just something CEOs talk about after their latest weary look at their ops headcount and legacy IT budget, when they’ve already concluded that there’s not much there that they can cut. GS have an additional self-imposed constraint: they try very hard, as a matter of longstanding corporate doctrine, not to respond to a downturn by mindlessly slashing headcount. They don’t like the IP loss and the inevitable rehiring when the cycle turns. They prefer a regular annual decimation, pour encourager les autres, and leave it at that. But they must still address costs, so they will always look towards tech investment when business is quiet.

In other words, this looking for a better, cheaper IT solution is yet another exercise in hope over experience. But unless Goldman has a super costly technology fiasco, it’s unlikely word of a failure will leak out; botched technology projects tend to be pretty well hidden. But it would manifest itself, if weak revenues persist, in eventual deeper headcount cuts, the very outcome Blankfein hopes to avoid.

Of course, distress for Wall Street producers, even if it is painful for them, is so mild compared to what ordinary people experience that the greater public won’t get much gratification from retrenchment among the financial classes. One can only hope that the new normal in finance also leads to newfound humility, but I would not hold my breath.

Print Friendly, PDF & Email


  1. Clive

    Oh Yves, this feature is so-ooo-o right on the money (literally). Unfortunately I’m travelling today so can’t elaborate (especially as I have to be careful to anonymise my anecdotals) but I can off the top of my head recount half a dozen failed IT project which burned $5M equivalent minimum each.

    Often an inadequate business process was attptedly sorted out with a tech fix. This is completely the wrong approach as you have to fix the business process first. Savings from IT are icing on the cake, if they exist at all.

    This piece is so good that it alone justifies a contribution for the NC Fundraising. I’ve known consultants charge $100’s of thousands for a lot less uspeful analysis.

    1. YankeeFrank

      Ditto Clive. In my years working IT at NYC investment banks and Wall Street “support” firms I’ve seen way more fiascos than successes. The quote you use is dead on: tactical solutions are favored (by smart IT managers) as strategic ones inevitably fail, and fail big.

      However, I will add that even tactical ones fail and remain unused quite often as well. IT managers deploy so many apps that never get used it reminds me of virtual space litter, though at least its virtual. They often get credit by checking a box marked “deployed”, regardless of whether a project works or is used by anyone at all.

      The quote regarding data — how much of what’s needed for automation doesn’t really exist — is very true, and I’d add that even the data that does exist is often extremely difficult to normalize and make usable. Bond data — for one example — is notoriously difficult to aggregate and normalize given the veritable sea of differing types and their vagaries. One can purchase such data for a tidy fee, but often there is a gap in what is covered, or the data doesn’t include some important fields, or different desks use data in different ways and need it normalized differently, or it is almost impossible to integrate with existing internal data stores — the exceptions become the norm and automation goes out the window.

      Another less technical part of the problem is that its extremely unsexy work that requires vast amounts of institutional knowledge and finance knowledge (not a strong suit amongst IT types). So no one wants such data jobs long term, yet that is the only way to make a strong data store. Most prefer working on high-visibility projects, not boring infrastructure, and many managers have no respect for “the bits that don’t show”, especially at the lower level investment banks.

      Also, there is often no funding for such projects as most IT funding comes from specific “desks” (and comes right off the top of their bonuses), and no one wants to pony up for projects that cost a fortune and will help others as much as themselves.

      Anyway, I could go on, but this post is spot on. Don’t believe anything Blankfein, et al. say about this stuff — if they could do it they would’ve done it already.

      1. Can't Help It

        Yep would agree with this. I think financial knowledge can be bridged. Now I did not actually work on the data side, but I took and passed all 3 levels of the CFA so that I could work better with the users. Not that difficult really as long as you put some hours into it.

  2. JGordon

    “the slightly chastened Masters of the Universe seem constitutionally unable to recognize that their own actions might have something to do with the fix they are in.”

    I was reading something recently on sciencedaily that said that people who get too much power tend to become clinically insane. And of course, people who want a lot of power start out being a least a little insane to begin with, so when they actually attain power that is something of a double whammy. That would explain not just Obama and Wall Street bankers very well, but also psychiatriasts who are prone lately to diagnosing anyone who doesn’t love Obama, bankers, or psychiatrists with “oppositional defiant disorder” or the like.

    So in other words, don’t hold your breath waiting for these “people” to recognize the error of their ways. All we can do is cross our fingers and hope that the collapse cleans out the whole mess of them.

      1. Melody

        A lefty talker yesterday described these uber-rich MOTU as having a mental illness–much like hoarders. Instead of amassing piles of newspapers, or napkins, or pizza boxes (that are just “too good” to throw away), they amass dollars. And like the hoarders, there are never enough digits in the bottom line of the bank account. Viewing these folks through the lens of mental illness might make us a bit more compassionate??? . . . uummmmmm, not so much!

        1. Max424

          “Viewing these folks through the lens of mental illness might make us a bit more compassionate???”

          Compassion? Interesting. Never thought about it. Do I feel sorry for Lloyd Blankfein? As a (fallen) Buddhist, I should (he’s deranged!), and I definitely would too, if Lloyd quit his job, said he was sorry, and proved it by swabbing latrines for Occupy Wall Street .

          Sun Tzu said, “Know thine enemy.”

          My way of thinking, since I came of age thirty years ago, is to view these people, always, thru the prism of insanity. It makes it easier to predict their future moves. In fact, it makes almost impossible not to know to certainty, where and when* they will strike next.

          They are stark raving nuts, clearly, and just as clearly, they have the power, meaning, they are capable of anything, including, as per their next Grande Plan, ending all life on this planet as we know it.

          So they must be stopped (the unhinged pursuit of superfluous and extra-meaningless wealth by our elites must end!), and best way to stop them is to get the word out to the general street population, that these people are not capitalist gods, to be esteemed, they are in reality, just plain old garden variety lunatics –only with money and power.

          *We know the “why.” They’re crazy.

        2. KV

          Joseph Heller, an important and funny writer now dead,
          and I were at a party given by a billionaire on Shelter Island.

          I said, “Joe, how does it make you feel to know that our host only yesterday may have made more money than your novel ‘Catch-22’ has earned in its entire history?”

          And Joe said, “I’ve got something he can never have.”

          And I said, “What on earth could that be, Joe?”

          And Joe said, “The knowledge that I’ve got enough.”

          –Kurt Vonnegut

        3. banger

          No, hoarders are the wrong comparison. Money people do so for the status–status-seeking is, perhaps, the greatest motivation for human activity. Because we have collectively defined status are relating to wealth then the strongest among us will gravitate towards making money. Those that practice morality, generally in our culture, do not have high status and the system is rigged to punish moral behavior–that is our culture. It is our decision whether or not we want to continue that way or seek another path.

          1. Mark P.

            Banger is right. It’s about the status and the power — money is tokens or dibs representing the full resources available to a human society.

  3. zygmuntFRAUDbernier

    Uncle Lloyd and Co. are still doing God’s work,
    I suppose …

    It seems to me doing God’s work is not
    always fun, judging by history: just look
    at Mother Teresa or Jesus. Worse still is
    the life of Job, who wrote the Book of Job.

    Uunless … but no. I will wait for any new
    Divinity-related messages from Uncle Lloyd.

  4. fresno dan

    “Lloyd Blankfein predictably makes “new regulations” the top scapegoat for Goldman’s reduced profit outlook, along with an uncertain market outlook.”

    Yeah, the old model of defrauding muppets by pawning off MBS’s, and than getting bailed out by the government which then led to record bonuses is obviously the best policy the government can pursue…for the top 0.0001%

    I can’t help but think of the other article posted here today about the Army Corps of Engineers and the “culture of impunity for failure…”

  5. charles sereno

    “That is part of why banks gravitate towards tactical solutions, not strategic ones.” (ex Goldman informant)

    Pyrrhus of Epirus (sans hair) looked amazingly like Lloyd Blankfein.

    Pyrrhus’ often denigrated “Pyrrhic victories” have gotten a bum rap. He was safely successful on balance, cutting losses, not taking needless risks. His strategy consisted in the pursuit of wise tactics — “not all eggs in one basket,” “win more than you lose.”
    This is also an old Goldman tradition. They survived the 1929 crash even though neck deep!

  6. JerseyRick

    There’s a lot more to Ops than touching the initial trades, too. I know of back offices with straight-through rates in the 90%s who still employ hundreds of clerks to research exceptions spit out from automated custodian reconciliation tools. What couldn’t easily be automated was moved to low cost sites. Years back that was NJ or NC or FLA, now it’s Southeast Asia. Fixing things is hard, moving work is easy and hits the bottom line fast.

    Clive has it right. I’ve seen systems built or vendor systems implemented that unthinkingly locked in the same broken processes and, surprisingly, didn’t deliver (well, by definition they always “succeed” if they have stakeholders with sufficient influence, their failures become manifest downstream).

  7. Can't Help It

    As someone who used to work at a major financial institution doing IT, there’s one really big cost that you can cut in theory and that is getting rid of mainframes. Those monolithical things cost a TON to maintain as opposed to getting off the shelf commodity hardware nowadays. Just compare the cost of IBM mainframe vs an HP Server.

    Another pipe dream that I have often heard of is this XMas wish of having just one single platform for fixed income another for equities, and another for derivatives. People not doing IT may wonder why this is so difficult, but it is really difficult. Vendors like Calypso doing all things and every thing in one platform exist but IMHO they suck. No hard feelings Calypso developers, but having business logic all over the place do not constitute a robust software!!

    1. YankeeFrank

      Indeed. In fact, Calypso is so unusable most firms have a cadre of coders dedicated to making front ends for it that compartmentalize and limit its functionality, making it somewhat usable for non-Calypso gurus.

    2. Susan the other

      I think this is why I have trouble with the saying “complex systems fail” or are inefficient, etc. Because when I hear this stuff about IT I see the problem as one created by totally simplistic solutions. If the complexity of GS’s software matched the complexity of its operations, then that would really be a complex system – and it probably would not fail. I would also imagine that the more complex an IT system is, the more it has been analyzed and understood and the easier it is to fix.

      1. Can't Help It

        Actually complex systems do fail more often at least in IT. The goal of software engineering is to manage complexity and often times that means simplifying things and reducing the number of moving components.

        I would however like to mention that many times a Wall Street project would fail not because it’s complex. Rather that the requirements are poorly specified. Traders, etc want a “great” system, but many times they would not even sit down and work things out with you. I once had to start building a system where the requirement was to compare “anything” with “anything” (anything here refers to some financial contract with no further definition). It’s the same as telling someone to just build a building.

        Thank the Lord I am out of the whole Financial IT hell hole.

  8. Matthew G. Saroff

    Well, hell. If you outsourced your brokers to India, and gave them a base salary of $15K/year base salary, and a $50K bonus, you would have tens of thousands of qualified brokers applying for jobs.

  9. Sleeper

    Here’s an alternate take –

    The contracts are often not designed to provide and actual workable product but rather to create an additional opportunity to loot the enterprise.

    Here’s how it works –

    The customer knowing over pays for a raw material or service and the supplier give back part of the cusuproduct

    1. Sleeper

      Here’s an alternate take –

      The contracts are often not designed to provide and actual workable product but rather to create an additional opportunity to loot the enterprise.

      Here’s how it works –

      The customer knowing over pays for a raw material or service and the supplier gives back part of the customer’s payment. This payment is of course fungible.

      The payment to the supplier is an expense and is written off on taxes as an expense.

      The whole transaction is based on the size of the giveback rather than product or service quality.

      This sort of transaction is commonly used in the purchase of plastic resin for automotive interiors, weather stripping,and yarn for textiles to name a few.

      Here are some warning signs that a “giveback” arrangement is in force.

      1) substandard supplies and services
      2) Higher than market pricing
      3) continued contracts with suppliers of failed products

      Oh and this is legal – just ask the IRS about givebacks.

  10. lambert strether

    If the IT systems were truly integrated, and if the data flows could be made transparent (modulo access privileges), would these systems perform better at looting, or worse? (I don’t mean merely stealing, a la Jon Corzine, I mean overcharging and people giving up because they can’t get justice from the call center).

  11. RepubAnon

    Maybe they can go to a cloud-based system. As this is the latest buzzword, success is assured!

    (Mordant chuckle).

  12. zygmuntFRAUDbernier

    If I type in Google:
    Systems Analyst new york NY jobs
    The second “hit” mentions 5223 job openings.

Comments are closed.