Banking StandardsWritten evidence from Finance Watch (V007)

Preliminary Remarks

Finance Watch is an independently funded public interest association dedicated to making finance work for the good of society. We are pleased to have this opportunity to contribute to the PCBS’s scrutiny of the Financial Services (Banking Reform) Bill on the question of whether to introduce a ban on proprietary trading within banking groups that contain a ring-fenced entity.

Finance Watch is an advocate of structural separation of commercial and investment banking activities. Whilst proprietary trading raises prudential concerns, we feel that these are best addressed by strengthening the ring-fence proposal. Separating proprietary trading is difficult and, for example, increases complexity and the possibility of loopholes. Therefore, on balance, our analysis argues against a ban on proprietary trading. In our view, trading activities should be located in well regulated, supervised, and capitalised investment banks, not least to facilitate visibility and control.

In short, the simpler the rules the stronger the ring-fence. The stronger the ring-fence the less need there is to ban proprietary trading.

Summary

Proprietary trading gives rise to general concerns, more specifically proprietary trading within banks gives rise to prudential concerns because of the possibility that losses will be borne by society, most directly via some form of bail-out.

Additionally, recent research from Germany has shown that the presence of proprietary trading activity alongside client-facing activity in banks creates the potential for conflicts of interest directly related to banking standards.

These concerns are to some extent addressed by ring-fencing; however compared to full structural separation there remain concerns that the ring-fence will not completely remove “perceived implicit guarantees” and their related subsidies.

A strong ring-fence is therefore the priority. A ban on proprietary trading could reduce risk but the difficulty in defining and enforcing such a ban and the likelihood that this would allow loopholes to undermine the overall robustness of the ring-fenced approach is a greater risk.

Separating proprietary trading from market making, or indeed hedging from speculation from arbitrage, is very difficult. For example both market making and hedging necessarily involve taking a view on price moves and expressing that view on your own account.

Proprietary trading is one among a number of investment banking activities that are best located and regulated within a well-capitalised investment bank structurally separate from commercial banking activities and subject to prudential controls. For example, the majority of market making faces only the financial sector. Less than 10% of OTC derivatives face non-financial firms (BIS), less than 15% of UK debt securities are issued by non-financial firms (ECB), world trade is less than 5% of total foreign exchange activity (WTO & BIS).

The difficulty of distinguishing between different types of trading activity and its overwhelmingly financial nature underline the need for a simply defined, robust ring-fence that will resist attempts to undermine it over time.

The UK and European banking sectors continue to rely on sovereign support, whether through direct ownership, central bank liquidity or other measures. Far from imposing a cost on banking and hence on society, we consider that structural separation, (including in the form of a strong, loophole-resistant ring-fence) is key to stabilising the sector and would help to restore the flow of credit between banks and from banks to the real economy.

Questions:

Stability

1. What prudential concerns does proprietary trading within banks give rise to? To what extent are any concerns being addressed through existing measures and proposals, including ring-fencing?

Proprietary trading within banks gives rise to prudential concerns because of the possibility that losses will be borne by society, most obviously via some form of bail-out.1 Own-account trading of any kind conducted within an entity which is partly or wholly backed by implicit or explicit state guarantees, such as banks, encourages asymmetric risk-taking (heads I win, tails the taxpayer loses). This contributes to micro- and macro-prudential risk and can increase resource misallocation and systemic risk.

These concerns are to some extent addressed by ring-fencing, in particular if the ring-fence effectively separates trading from those parts of the banks which must be bailed out and if the investment banking entity which bears trading losses is not too-big-to-fail. However compared to full structural separation (which Finance Watch advocates),2 prudential concerns remain. As the PCBS’s first report notes, ring-fences are fallible (see also our response to question 4 below). Given a leaky, badly maintained or weak ring-fence, there is a danger that losses from proprietary trading could cause problems for ring-fenced entities. Resolution tools, no matter how well designed, can only deal with so much. Therefore, there remains a danger that State intervention at taxpayers’ expense will be required and that “perceived implicit guarantees” will remain.

Furthermore, allowing some trading activities within the ring-fenced entity could increase prudential concerns. All own-account trading contains a proprietary element, including those which are labelled hedging and market making (see our response to question 7 below for more details).

We note that the government’s response to the PCBS report states that it “strongly agrees with the PCBS that the overall objective of banking reform should be to curtail any perceived implicit guarantees enjoyed by banks seen as ‘too big to fail’” (para 2.6).

The draft legislation published 4 February 2013 provides for some trading activities to be allowed inside the ring-fence in certain circumstances provided they do not threaten the continuity of core services (142D).

This gives rise to some concerns in part because proprietary elements are present in all trading activity, making it difficult to distinguish between trading activities on this basis. Allowing trading activities within the ring-fenced entity increases prudential risk, as noted above, but is also likely to weaken the ring-fence, as former ICB members Sir John Vickers and Martin Taylor have testified to the PCBS. Attempts to distinguish between different types of trading will require constant monitoring of loopholes that arise and give banks an incentive to “innovate” in order to benefit from explicit guarantees.

The same logic applies to any attempted ban on prop trading. As has been seen with the implementation of the Volcker Rule in the United States, distinguishing and separating proprietary trading can lead to complex rules. The danger exists that the more complex the rules the more likely banks will “innovate” to benefit from (in the proposed UK case) implicit guarantees created by weaknesses in the ring-fence. In addition, clarity is reduced for investors and regulators in this situation.

[NB—The general objective of the PRA as amended by section 1 of the 4 Feb draft bill, which can be summarised as ensuring the continuity of core services in the face of internal and external threats and of ring-fence bank failure, appears to us to be narrower than the three objectives set out by the ICB, which can be summarised as protecting taxpayers, ensuring the continuity of essential services and curtailing government guarantees (PCBS 130).

In particular there is no express reference to “mak[ing] it easier to sort out both ring-fenced banks and non-ring-fenced banks which get into trouble, without the provision of taxpayer-funded solvency support” or to “curtail[ing] government guarantees, reducing the risk to the public finances and making it less likely that banks will run excessive risks in the first place.” (PCBS 130, emphasis added)

Framing those two ICB goals as subsidiary parts of the continuity of core services objective, for example by reference outside the legislation to resolution and bank conduct (Govt. response para 2.15), is not enough in our view to ensure that the ICB’s public policy goals are fully taken into account by the PRA. We fear that the ICB’s objectives will not be fully translated into legislation without an explicit reference in the PRAs statutory objectives to reducing the role of taxpayers in bank failures and to curtailing government guarantees.

The public policy grounds for doing so are summed up clearly by the PCBS: “A guarantee, whether implicit or explicit, distorts incentives of managers and creditors, encouraging them to pursue excessive risk and leverage. It also distorts competition, and the allocation of resources, away from smaller banks to those large enough to be regarded as systemic” (PCBS 104). These objectives are central to the effectiveness of the secondary legislation envisaged in the bill and to the PRA’s statutory review of ring-fencing and should therefore be clarified to ensure that the purposes of the ring-fence as outlined by the ICB are not undermined or eroded (PCBS 171).]

Standards

2. To what extent does the presence of proprietary trading activity alongside client-facing activity in banks create the potential for conflicts of interest and affect remuneration practices, culture and banking standards?

Research from Germany into banks which managed assets for clients and undertook proprietary trading shows that:3

Banks systematically push stocks from their proprietary portfolio into their retail customers’ portfolios.

Those stocks systematically underperform compared to both other stocks in banks’ proprietary portfolio and other stocks in households’ portfolios.

Customers’ portfolio performance at banks with prop trading is significantly worse than at those without.

A strong ring-fence, or better still full structural separation, would go some way to mitigating this problem by separating commercial banking from the trading culture of investment banks. One way in which separation would help is if it gives investors a clearer choice between investing in commercial or investment banking and thereby going some way to insulating investors in retail banks from investment banking culture. For example, there is evidence that investors would rather that investments in retail banks were insulated from cases of miss-selling in investment banks.4 Steps to shift the focus of financial markets from short term price moves to long term investing would also help in this regard.

3. How adequately do current conduct rules manage any potential for conflicts and harm to standards? To what extent could stronger conduct rules address any problems?

4. Are there stronger grounds for concern about proprietary trading when it is conducted in a group which contains a ring-fenced bank than if it is conducted by a standalone wholesale bank? If so, why?

Yes. Proprietary trading within a group that contains a ring-fenced bank raises concerns that may not apply to proprietary trading in a standalone investment bank because funding and subsidy leakages may occur between the parts of a ring-fenced entity. As mentioned above and in the PCBS’s first report, fences may be leaky, may require maintenance and may be blown over in a crisis.

For example (a) if the ring-fenced and non-ring-fenced operations are funded with capital raised at the holding company level, the blended cost of that capital would necessarily include a subsidy; (2) if, in a crisis, the ring-fenced bank were able to transfer surplus capital and/or liquidity to the non-ring-fenced entity, that would also represent the extension of a subsidy to the proprietary trading operations; (3) if the market were to doubt the effectiveness of untested resolution mechanisms it might supply capital to the non-ring-fenced proprietary trading activities at artificially low rates; (4) in a worst-case scenario, a serious failure in one part of the group resulting from proprietary trading could result in a loss of market confidence in other entities in the group and require government support for one or more of them.

Full separation of well-regulated and well-capitalised, not-too-big-to-fail, investment banks would greatly reduce the reasons for a ban on proprietary trading. Such investment banks would be able to take more or less proprietary risk as they saw fit (and as their shareholders saw fit) without posing systemic risks. Likewise, within a group containing a ring-fenced entity the stronger the ring-fence the less the need for a ban (the strength of the ring-fence is also affected by its location, as is discussed more fully in response to question 7). From a regulatory and a reporting perspective such investment banks offer the best chance to regulate and report on a variety of investment banking activities including proprietary trading.

Bank lending and facilitating lending via securities markets are fundamentally different banking activities. Allowing these activities to mix freely has proved very costly for the economy and financial system in the past

Practical Considerations

5. How, if at all, would prohibiting proprietary trading by banking groups affect the case for implementing a ring-fence? Would there be any practical benefits from implementing both such measures at the same time?

A ban on proprietary trading does not undermine the case for ring-fencing. Structural separation remains a critical mitigant of prudential concerns. Any ban on proprietary activity would aim to support structural separation. However a ban on proprietary trading may not be the best way to strengthen the effectiveness of separation.

It is likely that for practical purposes only a small part of proprietary trading could be banned from groups containing a ring-fenced entity, ie dedicated proprietary trading activities (see our response to question 7 below). The remaining investment banking activity would still require a robust ring-fence.

Implementing a ban in addition to the ring-fence might remove a source of potential loss and asymmetric risk-taking from a ring-fenced group; however, the difficulties of regulating and enforcing such a ban are likely to be considerable, as has been seen in the United States. In our view, the benefits of eliminating a small amount of dedicated proprietary trading are likely to be outweighed by the strong likelihood that gaps elsewhere in the ring-fence would be systemically exploited. The complexity of defining and enforcing such a ban could undermine the overall robustness of the ring-fenced approach.

6. What powers does the regulator already have that could be used to prohibit banks from conducting proprietary trading? What are the constraints on any such powers?

7. What are the main challenges in defining proprietary trading, and how could these best be addressed?

Distinguishing and separating proprietary trading from other trading activities including from market making and from hedging, is very difficult. Trading is trading.

Market making necessarily involves taking a view on how prices are going to move, and necessarily involves holding an inventory. These are also the two essential elements of prop trading. All market making (as opposed to broking) involves an element of prop trading.

To make markets involves standing ready to buy and sell, and communicating to others in the market the prices at which you are willing to do so. Deciding on those prices requires deciding how happy, or not, you are to have the instrument in question on your book and adjusting your bid and ask prices accordingly. Market making does not involve simply adding a margin to a (hypothetical) mid-price, it involves adjusting your price to buy and your price to sell to “win” or “miss” client trades and to adjust your inventory accordingly, all the while bearing in mind your reputation.5

Similarly hedging necessarily involves transforming rather than eliminating risks. Simply put the only way to eliminate a risk is to sell the source of that risk. A hedge, by definition, is something other than selling the underlying source of risk. Therefore hedging always involves taking a proprietary view eg on counterparty risk, on the movements in an index vs. the price moves of a component, on movements in prices of similar instruments with different maturities and so on.6

Managing derivatives positions implies by definition opening risk positions on multiple parameters (market level, interest rates, volatility, correlation, etc…) and in the reality of a trading operation there is very little, if any, economic difference between “pure prop trading” and the management of client facing derivatives positions that require the bank to take a view on the evolution of those market parameters. This issue is particularly important in the UK case given the size of derivatives positions held by the largest UK banks.7

The fundamental similarity between all trading activities and their difference from commercial banking activities is one reason why we prefer full structural separation to ring-fencing; and, in the absence of full separation, a strong ring-fence that would keep all market making activities out of the subsidised ring-fenced entity.

The difficulty of distinguishing between trading activities provides a strong argument for simplicity, with regard to proprietary trading but also in the location of the ring-fence. Aside from the possibility of losses, a complex definition of permitted trading activities, either inside the ring-fenced entity or elsewhere in the group, will be subject to attempts at regulatory arbitrage and to degradation over time. Simple rules, on the other hand, would simplify the choices of investors in commercial and investment banks with their respective risk profiles.

Simple rules that group trading activities in an investment bank, separated from commercial banking activities, would aid regulatory control, including addressing public interest concerns across a range of investment banking activities. For example market making involves activities that might not be fully in the public interest. The “real” economy might need investment banking (to raise large scale finance) but, funding subsidies aside, one can reasonably ask how much investment banking needs the “real” economy. The majority of investment bank trading appears to be directed elsewhere. Less than 10% of OTC derivative notional outstanding faces non-financial firms.8 Less than 15% of UK securities outstanding (excluding shares) are issued by non-financial firms.9 World trade is less than 5% of foreign exchange activity.10 In short, the vast majority of financial market activity is not concerned with financing the “real” economy. Nevertheless it is banks, and only banks, that organise the issuance of these securities and make markets in them. It is not simply proprietary trading that should be insulated from public support, but also finance-to-finance securities and derivatives trading that should be forced to stand apart from (implicit) public support.

It is sometimes argued that separating all market making activity from ring-fenced subsidies would have costs as well as benefits. Specifically, it might increase market making costs for securities that are issued by non-financial firms (market making in corporate bonds), or instruments that are used by non-financial firms to hedge, eg interest rate and foreign exchange risks. We would contest this, above all we would note:

First, we would fundamentally contest the premise of the trade-off. While UK and European banking remains in dire shape (in public hands and often in receipt of central bank liquidity provision) measures to stabilise the sector and bring a return of confidence are more likely to decrease (stand-alone) funding costs than to increase them. Comparisons between a banking system post-structural separation and a theoretical healthy banking system pre-structural separation serve no purpose in assessing current policy options.

Second, as pointed out above only a small percentage of investment banking activity relates to non-financial firms. Removal of the funding subsidy (in conjunction with other measures) might remove some trading activity which does not relate to financing non-financial firms. This could be considered as a positive development for taxpayers as it would reduce both systemic risk and the “taxpayer put” subsidy for these activities.

Third, there is no reason that a loss of subsidy would necessarily be passed on to corporate customers. Given the right incentives banks could choose to reduce their margins rather than pass on any increased funding cost.11 Investment banks in recent years have been targeting very high returns on equity eg much higher than GDP growth.

Fourth, for essential activities like making markets in Government bonds, funding of inventory is not an issue as the assets traded (Government bonds and bills) can be easily financed through repurchase agreement (“repo”) operations which are, regardless of market circumstances, always the easiest and cheapest way of refinancing available for such activities.

Fifth, if one considers the total cost of funding of a bank (and therefore of a universal bank) is a direct function of the risks taken on the asset side of the balance sheet and if excluding a number of risky trading activities increases the cost of funding of these activities, this will also mechanically decrease the cost of funding of ring-fenced activities which, by definition will be positive for the economy and for society.

In short, there is great difficulty in distinguishing and separating proprietary trading from other trading. All own-account trading contains elements of proprietary trading, including market making and hedging of “simple derivatives” and other trading activities that the law as currently drafted might allow inside a ring-fenced entity. We feel the challenges of distinguishing and separating proprietary trading are best addressed by not attempting to do so—and instead by corralling all trading activity into well regulated, supervised and capitalised investment banks.

Overall Assessment

8. Given the factors above, how would you assess the case for:
(a) Including a prohibition on groups containing a ring-fenced bank from engaging in proprietary trading within the Banking Reform Bill

While proprietary trading in banks gives rise to prudential concerns, attempts to prohibit it also come with dangers. On balance, Finance Watch feels that rather than trying to ban prop trading, the best approach is to strengthen the ring-fence to ensure that losses from proprietary trading cannot blow it over.

Only the most obvious proprietary trading could be banned with ease, while the remaining own-account trading will retain proprietary elements. Moreover any attempt to define proprietary trading is likely to create opportunities for banks to “innovate” around the definition. Simple rules reduce the opportunity for lobbying12 and are easier for regulators to enforce.13 Simple rules also increase the transparency for investors. Finally there are also regulatory advantages from keeping proprietary trading in investment banks under the control of banking supervisors and regulators and capitalised under Basel and Capital Requirements Directive regulations.

Finance Watch has been active in the banking separation debate in several places in Europe (see our responses to the Liikanen Commission and Report, and to the French Parliament for example).14 We would note that the proposed approach in the UK is closer to that of the EU’s Liikanen Commission, which did not recommend a ban on proprietary trading, than current proposals in France and Germany are. In this regard, the UK has the opportunity to inspire a race to the top.

(b) Implementing or creating reserve powers for a proprietary trading ban through some other form

(c) Requiring that future reviews of the operation of the ring-fence have an explicit mandate to consider and report on the case for such a prohibition

(d) Using the Banking Reform Bill to give the regulator a reserve power to impose such a prohibition on individual banks if it concluded that it was necessary

About Finance Watch

Finance Watch is an independently funded public interest association dedicated to making finance work for the good of society.

Its mission is to strengthen the voice of society in the reform of European financial regulation by conducting advocacy and presenting public interest arguments to lawmakers and the public. Finance Watch’s members include consumer groups, housing associations, trade unions, NGOs, financial experts, academics and other civil society groups that collectively represent a large section of European citizens.

Finance Watch’s founding principles state that finance is essential for society in bringing capital to productive use in a transparent and sustainable manner, but that the legitimate pursuit of private interests by the financial industry should not be conducted to the detriment of society.

13 February 2013