• This blog has demonstrated that the City of London’s wholesale business will not be impacted by Brexit. Despite this, mainstream media is full of sensationalist articles prophesying terrible job losses in the City due to Brexit. Those articles almost invariably use research reports from respected, prestigious bodies as the source for their job loss forecasts.
  • Unfortunately, these research reports are inadequate, and contain either no reasoned explanation at all of how they arrive at their forecasts, or feeble reasoning which does not stand up to scrutiny. This blog will analyse and demolish each of those reports in turn.
  • This particular post is devoted to Oliver Wyman (OW)’s offering in 2016, which is the most commonly cited paper forecasting City job losses due to Brexit.
  • OW’s piece appears, superficially, to be detailed and well-informed. However most of the piece is padding: generic information with no relevance to the job loss forecast it makes. It does contain some arguments in support of its forecast, but these are vague and superficial, and based on a fundamentally mistaken understanding of how the City works.
  • Despite its thin, mediocre content, OW’s piece offers a fascinating insight into the smug, self-reinforcing way in which certain publications achieve the status of “expert reports” without displaying any expertise whatsoever, and how that fake expertise is given an unjustified appearance of solidity by being constantly cited in the media.
  • Anyone hungry for useful information on Brexit’s impact on the City will come away from Oliver Wyman’s piece like another Oliver, Oliver Twist, wanting “some more.” The articles which cite it are an echo chamber with nothing to echo. The bias in that echo chamber is something of which we’ve had quite enough.


Oliver, Oliver

This blog has demonstrated in some detail that very little of the business transacted by the City of London with the European Union requires membership of the European Single Market. Despite this, many reports in the mainstream media assert the contrary, and churn out one scare story after another claiming that tens of thousands of jobs will disappear from the City of London due to Brexit (here, for example).

These reports almost invariably use some kind of publication from a respected source, such as an accountancy firm or a think tank, as their pretext. Unfortunately, the most high profile and commonly cited pieces, such as those by Bruegel and EY (formerly Ernst & Young), contain no meaningful explanation of how they arrive at their Brexit job loss forecasts. The much-cited companion pieces of the pro-European and partly tax payer funded think tank Bruegel.org (8 February and 9 February 2017) are little more than finger in the air exercises, of which the authors themselves say: “it should be noted that our estimates … are for the purposes of illustration and debate, and not intended as a forecast” (8 February: p. 2). Similarly, the sensationalist 83 thousand job loss forecast by EY for the LSE Group, whose dramatic headline numbers were delivered, verbally, to the House of Commons finance committee with much fanfare, are hidden behind a mysterious pay wall (all these reports will be analysed in separate posts). A modern day Oliver Twist looking for any detailed explanations of the basis for these estimates might very well exclaim: “may I have some more?”

Among the pieces of research which have provided the media with the ammunition for their Brexit scare stories, there do exist a few which disclose some kind of calculation mechanism for the potential number of jobs they expect to be lost due to Brexit. One such is a report commissioned by TheCityUK, a high profile campaign group purportedly seeking to promote the City’s interests, and written by another Oliver, Oliver Wyman (OW), a management consultancy with a well regarded financial services practice (whose work I indeed sometimes come across when researching financial services stocks for investment). OW is part of the $41bn market capitalisation conglomerate Marsh & McLennan (or “Marsh Mack,” ticker MMC), which also includes insurance brokers Marsh, pension consultants and HR advisers Mercers and insurance analytics consultants Guy Carpenter. The report’s catchy title is “The Impact of the UK’s Exit From the EU on the UK-Based Financial Services Sector.” It is often cited to support the claim that “20%” of UK located banking activity is dependent on EU membership. I analyse the weakness of that claim and OW’s articulation of it in detail in a sister piece to this post. Here, we will look at the whole OW report. Does it really offer us “some more”? Or is it just another piece of thinly argued, paid-for Remainer propaganda? Let’s find out.


Oliver Wyman’s forecast

OW’s estimates of the impact of Brexit on the City comprise:

  • A low end loss forecast in which the UK strikes a deal with the EU and in which it retains the financial services passport and maintains regulatory equivalence with the EU;
  • A median scenario in which passport is lost and equivalence is not maintained, but only EU related activities move from the City to other locations as a consequence;
  • A high end loss scenario in which the loss of EU related business (e.g. Euro Swaps) results in a knock-on effect in which other, related, businesses (e.g. the other, non-Euro Swaps activities) move as well or are closed.

I have summarised OW’s estimates for each scenario in this table:

OW estimate table

Here are the relevant paragraphs from OW’s report (pps 2-3):

OW estimate

OW estimate 2

To put this in perspective:

  • The low end 3-4,000 is around the total number of jobs which UK-located banks such as HSBC, UBS, Citigroup and JP Morgan have said they would move;
  • The middle scenario’s 31-35,000 is equivalent to the thinly argued figure advanced by the Bruegel group;
  • The high end estimate is close to the 83,000 which was purportedly forecast in the mysterious EY report.

These are dramatic numbers. The middle and high end scenarios represent 5% and 12.5%, respectively, of all financial services employees working in London. The low end figure, which represents my own estimate in a no-deal Brexit scenario, is of a similar size to the jobs lost at Phones4u when it went bust, so no big deal.


Oliver on passporting and equivalence

The different scenarios on which the report’s range of forecasts is predicated are all based on the central assumption that securing financial services’ passporting rights and maintaining regulatory equivalence post-Brexit will have a big impact on the number of jobs the City of London can retain. It is therefore worth briefly defining those two things.

  • As described elsewhere, the financial services passport makes it illegal for regulators in any EU country to refuse authorisation to any financial services firm authorised in another EU country to operate in their country. But, as we argued, most of what the City does is sell wholesale financial services to other financial services companies, such as fund management services to UCITS funds in Dublin or Luxembourg, or Euro interest rate swaps to mortgage banks in Spain or the Netherlands. The financial passport is not required for such wholesale activity, but only to sell directly to retail customers (be they individuals or non-financial SMEs), which is not a significant part of what the City does.
  • Regulatory equivalence basically means adopting the same regulations and standards as those in force in the EU, such as MIFID. As we said in the previous bullet point, the wholesale nature of the City’s activities makes such regulatory equivalence unnecessary. Indeed, one of the benefits of Brexit is the ability it gives the City to free itself from some of the EU’s over-complicated or dysfunctional regulation. This would make the City more attractive and lead to jobs being moved to the City, not away from it.

OW’s report’s governing assumption – that the passport and equivalence are crucial – is therefore fundamentally at odds with our belief that they are trivial. The purpose of this blog post is to examine how the report uses those assumptions to arrive at its estimates of the impact of Brexit, and whether these are credible enough to satisfy a financially-minded Oliver Twist.


Summary of Oliver’s report

  1. Page 4 (the first page of the report following the executive summary) delineates the importance of the financial sector to the UK, with information on key headline data both for the industry overall, such as its contribution to the trade balance, and broken down by industry group (investment banking, sales and trading, insurance, exchanges etc.), such as revenues or contribution to GDP (GVA).
  2. Pages 5-6 state that OW estimates that “half of the sector’s revenues in the UK are from international and wholesale business,” meaning that half of the total, measured by activity, is wholesale and half of the total, measured by client location, is international. This 50% might seem low, as nearly all of the City’s activities are wholesale and most of that is international, but you must remember that OW is not just describing the City but all of financial services, which, in addition to the City, includes vanilla retail banking like the TSB or car insurance companies like Admiral. OW also, helpfully, break the UK’s financial imports and exports down by region, and the revenues of the main financial services activities (banking, insurance, asset management etc.) into domestic, EU and other international.
  3. Pages 7-8 discuss the network or cluster, or as OW put it “ecosystem effect,” in which different parts of the financial services ecosystem, such as equity fund management and equity broking, are mutually dependent on and reinforcing of each other. OW claim that it is because of that network effect that so much international financial activity is transacted in London, but that this also means that the removal of one part of that ecosystem might result in the disappearance of parts of the ecosystem which depend on it.
  4. Pages 8-11 describe the different regulatory outcomes, summarised above, in more detail, with helpful vignettes explaining various pieces of legislation.
  5. Pages 12-14, the meat of the paper, describe what OW believe will be the direct (“first order”) impacts on the bits of the City that trade with the EU and the indirect “ecosystem” (“second order”) effects – under the different regulatory scenarios outlined in pages 8-11.
  6. Pages 15-16 discuss the need for a transition period.
  7. The rest of the 22 page report consists of the conclusion and supplementary information such as definitions of terms used in the main body of the report.

Most of the report, surprisingly, is therefore fairly anodyne stuff which gives you a lot of general info, some of it useful, about the financial services sector. But it is only in pages 12-14 that OW offers any kind of justification for its estimates, and it is these that we shall discuss in detail.


Does Oliver understand the financial services “ecosystem”?

Before doing so, it is worth looking at OW’s discussion of the ecosystem effect, which suffers from some important and telling flaws. While its illustration of an ecosystem on page 7 (figure 4) does indeed show different financial services activities which work together as an ecosystem, the example is poorly chosen because there is no need for those activities to be located in the same place:

ecosystem 4

  • The diagram places the building society inside the ecosystem. But mortgage banks from all over Europe, indeed all over the world, buy swaps from investment banks in London which those investment banks trade over the counter (OTC) through inter-dealer brokers (IDBs) in London. That’s why finance is international and why London is so big. It is also why the EU is so exercised about London being the center for Euro Swaps trading.
  • The diagram shows the London asset manager buying a covered bond from a London investment bank. But the reason why London is so good at fund management is that it is home to investment specialists able to find attractive investments all over the world, from Colombo to Bogota. And the reason it is so good at investment banking is that its investment banks are able to source investment ideas for asset managers all over the world, from Hong Kong to Chicago. There is no network effect, at a time when communication has never been faster, easier and cheaper, for the fund manager to have the same post code as the investment bank.
  • Pension funds and insurers all over the world invest their money with asset managers in the UK – and all over the world.

This diagram is very strange. It represents a world which isn’t international, and is at odds with the incredible internationalisation and freedom of financial markets today. It also neglects an important point. All of these actors are financial institutions, not individuals or non-financial companies. It is because of this that financial markets are so international: with a few minor exceptions, financial institutions are able to use counterparties anywhere in the world – as long as the country doesn’t have capital controls.


How does Oliver arrive at his forecast?

For OW’s estimate to have any validity, it needs to make a case for the so-called “first order” effects, in other words for the direct impact of a loss of trade with the EU on specific parts of the financial industry in London. If none of these first order effects are valid, then the second order effects won’t exist either. Incredibly, the section on first order effects consists of three very short paragraphs on three industries: sales and trading, asset management and speciality insurance.

sales and trading

For all its purported expertise in financial services, OW is guilty of extreme sloppiness here. Just who is the UK based firm’s “client”? If it is José Savon, an acme French individual trading in stocks, then, yes, the City won’t be able to trade with such an individual without an EU financial services passport. That would, indeed, be the sort of retail transaction which requires authorisation in the EU which will no longer be facilitated by the EU financial services passport post-Brexit. But that is not what the City does. The overwhelming majority of sales and trading in the City for EU clients is carried out for investment managers or banks and other financial institutions, buying and selling various investments (equities, bonds, commodities, derivatives etc.). They might do so for their funds (in the case of investment managers) or to hedge their interest rate exposure (in the case of mortgage banks). The City’s “client” is, almost invariably, another financial institution. Such financial institutions can already, as things stand, transact in financial centers all over the world. French fund manager Carmignac Géstion can, as things stand, buy equities in Singapore or corporate bonds in Hong Kong. Landesbank Hessen-Thüringen can, as things stand, hedge its exposure by entering Swap contracts with banks in Japan.

It is such EU financial institutions, typically a retail bank, which look after José Savon and the other individuals (Hans Seif or Giuseppe Sapone) taking a flutter on the market, not the UK located investment bank which only deals with the EU retail bank. José Savon will typically place his order with a local online broker who then goes through a London investment bank to execute the trade. Such direct trading with retail consumers (whether individual or corporate), is typically carried out at the local level by local players; there are very few pan-European retail financial firms. In fact, one of the most well known, ING, has severely retrenched in favor of local players. One of the most successful currently is IG Group, and, as we have written elsewhere, it still has to obtain authorisation from local regulators such as the Banque de France to offer Contract for Difference (CFD) products in France.

It is very remiss of OW not to acknowledge this distinction. Their neglect could lead to an inaccurate calibration of the number of jobs lost due to Brexit. There are many people employed in sales and trading in the City, so if you wrongly assume that all of those serving EU clients in sales and trading will need to be transferred to an EU location because of Brexit, then your estimate of Brexit-related job losses in sales and trading will be very high. But, as we have explained, it is only the retail sales and trading jobs which would need to move, and these are relatively insignificant in number (we try to estimate that number here). Only by accurately counting the number of EU related retail sales and trading jobs in the City can a robust estimate be formed of the number of jobs that might be lost there “because Brexit.” By neglecting this distinction, OW simply side-step that vital calculation.

asset management

OW deserve credit here, at least, for their understanding of the fund management industry. As we explained elsewhere, the funds in which savers invest their money are independent entities, with their own board of directors, custodians and administrators. It is these funds which take the money from the investor when she buys units in them. In order for a fund to be sold to investors in the EU, that fund must be conform to EU regulations (UCITS being the most common EU-authorised vehicle). That fund can then, as OW puts it, “delegate” the management of the assets in the fund, i.e. the decision regarding which equities, bonds or other assets to buy or sell on behalf of the fund, to another company – this is the fund management company. It is the fund that has to be authorised in the EU, not the fund management company. This is the relevant text of the UCITS legislation (p 49):

delegated management

Article 13 of the EU’s UCITS legislation states unambiguously, in black and white, that third country fund managers can act as delegated managers of UCITS funds. UK fund managers will therefore be able to act as delegated managers of UCITS funds when the UK becomes a third country post-Brexit. As we explained, most funds sold in the EU are UCITS funds, issued in Dublin and Luxembourg, and virtually all of the City’s EU related income from asset management is from just this “delegated management” of such UCITS funds. The income from issuing such funds was already generated in Dublin and Luxembourg, not London. The income from the delegated management of such funds can continue to be earned in the UK when it becomes a third country, post Brexit. So nothing changes for the City post Brexit. In fact, UCITS funds provide the City with the perfect “Trojan horse” into the European fund management market.

However, there are two small fields of activity which might be impacted:

  • Distribution of funds to investors is a retail activity, and that is typically done at a local level. It requires a local brand and field agents who speak the local language. That’s how direct to consumer works. However, if there were any retail fund sales people based in the City and serving EU retail clients, then they would need to move (I describe how such retail fund sales jobs form part of my estimated 4,000 City job losses due to Brexit here).
  • Custody is also an issue. MIFID II allows the custodian to be located outside of the EU with the client’s permission. It seems however that under UCITS V, custody and fund administration services will have to be provided in the same country as the fund is incorporated (see §25). If that is the case then any custody work performed in the UK for UCITS funds, for example by State Street in Edinburgh or Glasgow, might have to move. However, that would have been the case whether the UK was part of the EU or not, as, under UCITS V, the custody will have to be performed not just in the EU, but in the same jurisdiction as the fund (typically Eire, Luxembourg or Malta). Whatever the case, just like retail distribution of funds, custody of UCITS funds is a minor activity in the UK; most of this is done out of Dublin or Luxembourg already.

Unfortunately, OW’s report does not go into this level of detail in specifying which precise activities might be jeopardised by the lack of a deal. This is important, because to correctly estimate the impact of Brexit on jobs in this area you would need to perform a minute and targeted count of the number of people employed in the UK to sell investment funds direct to EU retail clients, or of people in the UK working in custody teams servicing UCITS funds.

corporate and specialty insurance

Insurance is a largely domestic business. Sales, loss adjustment, customer services and capital requirement compliance are all performed locally. There are very few UK based insurance brokers who service European clients directly from the UK. There are many pan-European insurance operations, notably Allianz, Axa, Aviva and ING, but these all already have substantial domestic European subsidiaries. Aviva does not need to lose employees in the UK so its Belgian subsidiary can keep selling car insurance to Belgian drivers. The insurance market Lloyd’s of London underwrites risk for insurers all over the world, including Europe. Unlike most insurers, Lloyds’ business is therefore cross-border rather than domestic. It therefore provides the best available litmus test of whether OW is right about the impact of Brexit on speciality insurance. Its CEO, Inga Beale, announced on BBC Radio 4’s Today Programme that the market was opening a subsidiary in Brussels on 30 March 2017. Much to the chagrin of Dominic O’Connell, the BBC journalist interviewing her after the decision was announced, she answered his question – “how many people will go to Brussels, do you think?” – by saying

It will be minimal. We’re expecting Lloyd’s to have people numbering in the tens, ah, in the London insurance market at the moment there’s about 34 thousand people, so it’s really minimal impact.

Admittedly, this announcement came out after OW’s report, but it confirms what should have been obvious to its authors at time of writing. Ironically, the pre-eminence of national regulators in insurance was attested in 2016 by no less an authority than the Bruegel Group’s cherubic Dirk Schoenmaker (2016 Annual Report, p 53):


And that’s it. Three tiny paragraphs on three industries, with extremely thin, cursory arguments as to why they might need to move post-Brexit, and, above all, no detailed analysis whatsoever of the particular sub-segments which might be affected.


Oliver’s second order impacts

The second order impacts listed on p 13 are, surprisingly, a repetition of the examples offered on page 8 as “examples of ecosystem impacts” (figure 5):


The fact that OW basically repeated the same set of points almost verbatim in two different sections of the report might seem like a merely formal, incidental feature. But it is telling. As we have seen, the report stretches to a mere 16 pages, including two for the introduction and one for the conclusion. This of itself makes it a short piece, which has already been padded out with a lot of generic information about the industry and various pieces of legislation. The re-cycling of arguments in OW’s discussion of second order effects is notable because it means that the bits of the report where OW are actually making their substantive points have themselves been padded out. This is particularly significant in light of just how cursory OW’s discussion of the first round effects was, as we examined above. The padding constituted by this recycling of the examples of second round effects sits on top of very meagre arguments about the first round effects.

Turning to OW’s discussion of the second round effects, some, such as sales and trading, might be legitimate if activities like Euro swaps trading moved from London. As we demonstrated above, since most of the sales and trading in London is wholesale in nature only a trivial amount of marginal, retail activity will be forced to move. On top of that, the actual argument given by OW doesn’t hold water (bad pun intended). Liquidity is not concentrated in trading desks as such, but in internalised liquidity pools or on exchanges. These liquidity pools can “cross” client orders wherever individual trading desks are located. Moreover, liquidity and depth of markets matters on an instrument by instrument basis. If you have a market that trades $1bn a day that might seem very liquid, but if it’s 100,000 different stock lines each trading $10,000 in volume that’s terrible liquidity in the average stock. Contrariwise, if you have internalised flow of $80m every day in Banco Santander ordinary shares across your five sales and trading operations in London and Europe, that’s great liquidity and your customers will get a good deal, even if your flow in other stocks is mediocre.

Speciality insurance, as we saw, is also barely affected at the first order level, so is not going to suffer second order effects.

Other examples are flawed because of the international, wholesale nature of financial services, discussed above. Asset managers use brokers all over the world, and indeed certain big fund management centers, like Boston or Edinburgh, have minimal broker presence, and make a virtue of this fact, with some managers boasting that it removes them from the hubbub of broker chatter. Clearing houses can and do clear trades across the globe, as this extract from LCH Clearnet’s annual report (2015) makes abundantly clear:


The strangest one is financial technology. Why should a financial technology company in the UK need to locate its headquarters or programmers next to its customers in Europe? Software consists of lines of code, the most easily exportable product in the world. Financial technology such as Bloomberg terminals or SS&C reporting software are sold all over the globe. The world in which you need to be next to fund managers to create fund manager compliance software doesn’t exist.


Oliver’s numbers

Well then – we’ve got equally thin second order impacts to add to our thin first order impacts. Still, as we – not OW – established, there really are a few pockets of activity which are affected by Brexit. With much toil, in the companion piece to this post, we arrived at a figure of around 4,000 for the banking sector, in which the retail activities affected by Brexit are concentrated. If we can do that in our spare time, purely based on publicly available information, OW, with its highly paid full time staff and access to in-depth information not available to the public and guidance from company management at various levels, should be able to come up with a more precise forecast, based on the number of people working in each segment and how many of those would need to move. Indeed, calculating this specific impact would really be a service to the public, since these numbers are either not publicly available or very time consuming to extract. OW, by contacting each individual company and aggregating the results, cross-checking this with companies like recruitment agencies who have a good feeling for the number of employees involved, or with businesses who are customers of those industries or for whom those industries are customers, could provide meaningful numbers which it would be impossible or cost considerable effort for the layman to produce.

Instead, this is what they come up with:

OW estimate substance

This isn’t a typo. It is from page 13 of the report, not the nearly identical summary presentation of the numbers in the introduction, quoted above. There is no further analysis: no explanation of which particular activities, within each broad area of activity, might be affected, or of how many people work in them, or of how many of those might have to move. Nothing. Rien. Nichts. Niente. OW doesn’t even break its job loss estimate down by industry, which you might expect given that it had discussed the different industries affected (albeit, as we saw, in a highly superficial way). After thirteen pages you are back to where you started. On the way you may have absorbed some anodyne, possibly useful, generic information. On two of those pages, OW’s report may have offered a set of broad brush, superficial and minimal arguments regarding how Brexit might impact various broad areas of City activity, as well as a set of numbers to quantify that impact. But not a syllable as to how you get from one to the other. No real analysis as to why the loss of passporting or abandon of equivalence might result in any particular job loss number.

Any reader of Oliver Wyman’s report might well be entitled to say, like Oliver Twist, “please Sir, may I have some more?”


Padding in the service of ideology

The poverty of the analysis helps to explain the seemingly incidental detail, discussed above, of OW’s recycling of one of its arguments, in other words it helps explain the padding with which it has bulked up its otherwise thin argumentation. Without such padding, the weakness of the argument would be cruelly exposed; it would stand naked in front of the reader. The padding not only disguises the meagerness of the report’s arguments but attempts to lend weight and respectability to it.

What sort of effect will this padding have on a casual reader, or one unfamiliar with the mechanics of how the City operates? The generic information the report offers about the economic impact of the City’s various activities, or the geographical breakdown of its customers, or the various legislative instruments at stake in Brexit, may be bland stuff, available from many other sources. But, to the layperson, such a slickly presented slew of numbers and legislative précis is likely to give the impression of expertise, to make it seem that OW “really know their stuff.” If they do know their stuff, then the reader is psychologically predisposed to believing what they say, including, crucially, their job loss forecast numbers.

But, logically, there is no reason to believe someone’s job loss forecast for an industry because they can assemble trade or GVA statistics, or offer neat summaries of technical directives. The two abilities are distinct, and one activity (forecasting) is much harder to perform than the other. Given enough time and resources, anyone can put historical economic data and summaries of existing legislation together. But the future is unknowable, and the track record of even the smartest people in predicting it is very poor (as notably discussed in the work of Philip Tetlock, Dan Gardner and Nassim Taleb).

OW is using this generic information to create what Kahneman and Tversky describe as a “halo effect.” The term was coined by Edward Thorndyke to describe people’s intuitive predisposition to trust statements by good looking people: their good looks provide them with a halo which confers trust on what they say, even if it’s about a technical subject, like insurance, which being good looking doesn’t help you understand. Kahneman and Tversky extend the term beyond good looks to a whole battery of qualities, for example decisiveness, speed of thought or general knowledge, which are easy to perceive in a person and appreciated. Kahneman and Tversky observed that when we evaluate someone’s ability to perform something difficult to do – and whose difficulty makes it difficult to evaluate how good people are at doing it – like forecasting, we tend to judge them according to the halo effect imported from these other, more readily appreciable abilities (such as making neat-looking infographics).

In this case, the reader is not only confronted with a difficult question: “how many jobs will be lost in activity x, y or z post-Brexit in scenario a or b”? but with the even more difficult question: “how qualified is the person I am reading to answer this difficult question I struggle to answer?”. The generic padding acts as a halo in this case: it makes the authors seem to possess impressive knowledge, even if we lack the ability to properly evaluate whether they truly have any forecasting ability. The layperson reading the report, when confronted with this knowledge, is likely to think something like “this is complicated, I’ll struggle to understand it, they seem to know a lot about it, so I’ll trust what they say.”

One also needs to consider that most people won’t read the report in detail. Indeed I strongly suspect that not one of the journalists or politicians who cite the report have read any more than its executive summary, and that most have merely seized on the headline job loss numbers. So many won’t even be aware that there is padding, or of the nature of that padding, or of the shallowness of the underlying argument. But every page of padding they skim will give them the impression that there was something there which they could have learned from, had they read in detail. For such a reader, the padding actually creates an underlying assumption that the arguments supporting the forecast are there, somewhere, in the report – had they been bothered to read it properly. What the cursory reader (which is most readers) will see is a lot of seemingly well presented tables and diagrams, giving an impression of coherence and precision. The way most people read reports on unfamiliar and technical subjects contributes to the halo effect, prompting them to think that the job loss estimate must be as accurate as the tables and summaries are impressive – because they haven’t read the bits explaining the forecast in the first place.

Indeed, the padding might have a bludgeoning effect which helps the report make its point. The reader, when confronted with the generic stuff, could be excused for being bored. After all, none of this really explains why job losses should arise in the City due to Brexit, which is why she is reading the report. The more boring something is, the more effort you need in order to plow through it. Analysing a problem like the impact of Brexit on the City also requires effort. So by the time the reader gets to page thirteen his resources may have been depleted, making it harder for him to identify the weaknesses in OW’s argument.


Establishment Oliver

The poor quality of this report is shocking, but that is only half the story. The fact that a well regarded firm like OW could put its name to such a report is also worth investigating. Why would they potentially compromise their reputation with such a flimsy piece of research? One reason may be OW’s membership of the corporate elite. OW is a big, prestigious company in the highly aspirational management consultancy profession, which makes its money from advising big blue chip corporates (this interesting BBC radio programme on high achievers interviews the kind of management consultants who might work for a firm like OW here). A management consultant at OW is paid a six figure salary and will find herself making presentations to the boards of Footsie 100 companies. OW is therefore a part of the establishment and elite against which the Brexit vote was, in part, a protest. So OW’s authorship of the paper is clearly, if nothing else, an example of a certain type of establishment support for a negative view of Brexit on which we will comment in another post.

If you express a view about something very politically contentious, like Brexit, based on rigorous objective arguments, then you can argue that you are taking the role of a dispassionate, neutral, “expert” contributor to the debate. But we have seen that OW hasn’t provided any real justification for its very negative forecast about Brexit. This means that it cannot legitimately make any claim to neutral objectivity. Without such objectivity, the statement made by OW’s forecast inevitably becomes a partisan one, a political one. And this negative view of Brexit, which we can now understand as political, is entirely consistent with the pro-EU view of the professional establishment. In short, far from constituting an expert, technical forecast, OW’s research paper is the expression of the political views of the professional establishment of which OW is a core member.


Seal of approval

OW’s membership of the professional establishment doesn’t just contribute to this particular political bias, but also strengthens OW’s ability to present that bias as convincing and authoritative. As a reputable management consultant, OW’s name carries weight. The board of the Footsie company can use a recommendation from OW in support of a decision it takes – “as recommended by Oliver Wyman.” And this lends further credibility to OW as an institution, allowing it to claim that it has “advised the boards of [pick a number] Footsie companies in [pick a bigger number] crucial decisions.” This mutually reinforcing prestige and image of authority, when it is used to merely lend credibility to political points, rather than offer technical information on a particular subject, is, of course, a classic halo effect. OW and its clients are prestigious due to their position in the economic food chain. OW makes the clients look respectable as recipients of its advice, i.e. lends them a halo, and its clients make OW look respectable by retaining it as an adviser, lending a halo in return.

Because other purported experts will also be advancing the same political idea (and, like OW, disguising it as objective analysis), OW’s credibility will gain from singing from the same hymn sheet as other “experts.”But, in this particular case, the halo is detached from any substance. And that thoroughly unjustified halo is transferred to the very political and entirely mistaken view that Brexit will cause lots of jobs to be lost in the City.

By endorsing and lending its halo, as such a reputable entity, to a negative view of Brexit, OW contributes to the view that experts in general think Brexit is a bad idea. For laypeople who lack the specific knowledge required to understand such a specialised subject, a natural reaction is to turn to those they perceive as experts for help. OW’s misleading cultivation of the notion that expert opinion views the consequences of Brexit as negative therefore works to undermine the lay person’s confidence in Brexit. It is effective anti-Brexit propaganda.

In summary, OW’s research is an excellent example of why there is a mistrust of experts in general and connected to Brexit in particular. Many intuitively suspect that the “experts'” expertise is either phony, or enlisted to make claims in areas to which that expertise does not stretch. OW’s piece is a case in point. And many further suspect that this phony or tenuous use of expertise is in fact a disguised way to lend credibility to a particular political view, in which the “experts” often have a vested interest. OW’s piece does precisely that.


Oliver’s radiating halo

For all its superficiality, OW’s report has had a lot of influence. It is cited throughout the House of Lords report on the impact of Brexit on the City (pps 5, 12-13, 16 and 37), and, crucially, its estimates are taken as reference points without any kind of questioning. That report is frequently cited, among others by Reuters, Guardian, Independent and Daily Mail. Crucially, all those citing the House of Lords follow it in accepting the OW estimates without question.

Even the Eurosceptic Open Europe think tank in its own, more measured piece on the impact of Brexit on UK financial services, frequently adopts OW’s assumptions, as it does here (p 19; this £23-£27bn estimate is the subject of the sister piece to this post):

Open Europe 23 to 27

The report is also cited on eight occasions (pps 1, 4, 6-7) in an article by former Bulgarian finance minister and LSE academic Simeon Djankov, “The City of London After Brexit,” for the influential Peterson Institute (which employs Bruegel founder Nicolas Véron, whom Djankov cites, along with his Bruegel co-authors, in his piece). Like the House of Lords report, Djankov, showing poor form for an academic, simply parrots OW’s assumptions. The Financial Times’ European Editor Tony Barber, on 23 March 2017, praised an updated version of the paper, with identical references to OW’s piece, in these terms:

For a lucid assessment of Brexit’s impact on the City of London, look no further than this discussion paper, written by Simeon Djankov. Having served with distinction as Bulgaria’s finance minister from 2009 to 2013, Mr Djankov is director of the financial markets group at the London School of Economics.

His paper is timely because, nine months after Britain’s vote to leave the EU, it is clear that most major financial groups are implementing plans to move some jobs and business operations from London. Still unanswered questions include how far this trend will go, and which rival financial centres will benefit the most (my emphasis).

Djankov’s awful, shambolic paper will be dealt with in detail elsewhere, but for the purposes of this discussion we can restrict ourselves to observing that it restricts itself to repeating OW’s estimates as if they were the gospel truth. His conclusions are celebrated and puffed up in this piece by an FT editor, no less. Talk about halo effect.

So the halo effect used to lend credibility to this thin piece of research has worked beautifully, shining ever more brightly due to the addition of the many halos from its citation in influential publications.

When I first embarked on the project of researching the potential impact of Brexit on the City, I was hoping to find information and arguments that might help me understand which areas were at risk in the work of OW and its peers. Of course, such material would have challenged me to come up with counterarguments, supported by information I gathered, and potentially undermined my hypothesis. But I would have preferred it that way; I would have learned something and either changed my hypothesis or been more secure in it, having submitted it to serious challenge. Unfortunately, there is nothing of substance in OW and the work of the other prestigious groups or authors referred to in this post. Rather than contribute anything useful to the debate, they just cite each other in a smug game of “pass the halo.” Not only is it an echo chamber, it’s an echo chamber with nothing to echo.


Who pays Oliver the piper?

The report was published in 2016, after the Brexit vote, when the government’s Brexit policy was not finalised. As we saw, its estimates will make the average person worry that the City, and therefore the country, will suffer greatly if the UK doesn’t retain its passport or maintain regulatory equivalence post-Brexit. Not only are politicians affected by the fears of their constituents, most of them understand little more about finance than those constituents and can themselves easily be scared by reports like OW’s. The report therefore seems to have been designed in part to influence the government’s approach to Brexit, by encouraging it to make concessions to the EU in order to retain the financial services passport, or to retain regulatory equivalence with the EU after Brexit.

Why would OW want to do this? One reason is that such a view forms part of an ideological consensus among establishment professionals, as discussed above. However OW is not going to allocate expensive consultants and put its name to such a ropey estimate merely as a public service to its establishment chums. It has a business to run. The main reason why OW put its name above these silly estimates was because it was paid to do so. We therefore need to look at what motivated TheCityUK to pay for the report.

TheCityUK purports, as we have said, to be an advocacy group for the City. Unfortunately, its annual return on Companies House does not provide a list of shareholders, and its annual report does not disclose which members made significant contributions to its £5.1m turnover, or, more importantly, funded some of its research projects directly. So we don’t know who is pulling the strings there. We can, however, in the absence of hard information, venture a few hypotheses. The larger international banks have more income to donate to an organisation like TheCityUK than smaller City firms. Having a larger spread of activities, these larger banks will also have more “touch points” in connection with which they might want to influence the government. In the absence of evidence to the contrary, you would expect, as your “Bayesian prior,” that TheCityUK’s main funders will be the larger investment banks.

If we look at the CityUK’s stance, we can see that it opposed Brexit before the referendum. Its post referendum stance advocates retaining maximal ties to Europe, including regulatory integration, as we have seen in this OW piece. As we will discuss elsewhere, it is the large investment banks which have the greatest interest in maintaining the passport and regulatory equivalence. Unlike most companies in the City, the larger investment banks have, in addition to the typical City wholesale activities, a significant retail, direct to consumer, activity (which is described in more detail here). Indeed, direct access to EU retail clients by wholesale investment banks is a form of vertical integration which strengthens the position of the big banks versus their competitors. Independent fund management groups for example need independent, usually local, distributors or introducers for their products. If the big banks gain share in the retail market, they can use their role as adviser to their retail customers to promote their own fund management products rather than those of their independent fund manager competitors. Such an activity is marginal to the City as a whole, but significant for those big banks. It is such an activity which would need to be relocated in the event of Brexit.

Moreover, EU regulations, such as MIFID, are very burdensome, raising the fixed costs and complexity of operating in the financial industry. This makes it harder for small players to operate, and therefore represents a competitive “moat” for the larger players, who are more able to absorb the cost and manage the complexity. The EU’s regulations therefore, although obstructive for the industry overall, represent a relative competitive advantage for the larger banks.

In conclusion, OW’s weakly argued piece may be part of an echo chamber with nothing to echo, but that echo chamber clearly serves both the political agenda of the anti-Brexit establishment and the corporate interests of the large global investment banks.

Thanks Oliver, I think we’ve had our fill.