In the first blog post from The Flipper our new anonymous non de plume open to all across the industry to write under the cloak of anonymity, we look at the impact of CRD IV's capital requirements rules on the prop trading market.
Much is known about the challenges that banks are facing in complying with the capital requirements set out under CRD IV. However, less well understood is the potentially devastating impact on market markers and proprietary trading firms.
CRD IV and the associated regulation CRR are aimed at:
(a) minimising the negative effects of firms failing by ensuring that firms hold enough financial resources to cover the risk associated with their business
(b) increasing the likelihood that a firms' personnel take an interest in the longer-term success of the business rather than short-term gains (for which read ‘increased risks’).
Most proprietary trading firms, which are Part 4 FSMA authorised by the FCA, are affected.
There are a number of ‘simple’ rules which apply to remuneration of certain employees, however, much more contentious are the rules which deal with firms’ capital adequacy.
The rules provide standard models for firms to calculate their capital adequacy requirements, the upshot of which is that for futures (where the capital requirements can be spectacular) firms must set aside Pillar 1 position risk capital of an order of magnitude related to the notional value of the derivative’s underlying instrument.
To be clear, for short-end sterling interest rate futures the notional value is £1m.
Inserting this multiplier into the standard position risk calculations (CRR 326 onwards and particularly CRR 339) is tantamount to basing regulatory capital requirements off the risk of a total loss; ie. that the future goes to zero.
Admittedly the spectacular position risk figures generated are subject to various haircuts as firms determine their own funds requirements (e.g. CRR 92).
But still, a simple exchange traded and highly liquid fixed income or interest rate future requires regulatory capital orders of magnitude larger than any firm or exchange risk model would predict or clearer would require as margin.
In short, firms may potentially face a catastrophic capital adequacy hit, especially when trading futures at the short end of the curve.
Options are treated similarly to futures although the amount of the capital required to cover an option will depend on its delta.
Equities receive a more favourable treatment - the concept of notional value is of course not pertinent to equities so the regulators have taken a different approach that produces significantly lower capital adequacy requirements.
In contrast to the standardised approach, firms might try to follow the banks by running their own risk models to determine Pillar 1 position risk requirements.
However, agreeing a bespoke risk model with the FCA could take months, if not a year, and in the meantime firms would need to apply the standard model.
The good news is that even if a firm uses the standard model, there are a number of ways in the model pursuant to which firms can apply haircuts and netting to reduce their capital adequacy requirements. The bad news is it does not reduce the amounts substantially.
The other good news is that the FCA does seem currently open to considering applying different methodologies and calculations, although the timetable for (and results of) that are uncertain.
What the FCA has been spending more time doing of late is complaining that the Dutch have permitted firms which the AFM regulates to apply different models which lead to those firms having substantially lower capital adequacy requirements (and therefore a significant commercial advantage).
According to various people, the AFM has decided that:
(a) it will not permit any firms not currently authorised and regulated by it to come to the Netherlands
(b) it will move to the standard model some time in the next couple of years (probably on the same timetable as MiFID2 implementation).
So all EU firms will be in the same boat eventually, although this isn’t really a ‘win’.
Some argue (and with good reason) that proprietary trading firms (which generally have good pre- and post-trade systems and controls and internal risk management and substantial margin and net liq requirements) do not present a systemic risk of anything like the magnitude that should require them to hold capital of such enormous amounts.
This could (and will) drive some players out of the market, leading to lower liquidity, wider spreads etc.
Those firms which have been around (or have been subject to Pillar 1 CRR position risk requirements) for some time may well have long-standing internal risk models which have been agreed with the FCA.
Those firms which became Part 4 FSMA authorised because of the German HFT law or which are planning to become authorised and regulated under MiFID2 do not have that luxury.
Counter-intuitively, also, since banks most likely have long-established internal risk models, even though they are much more likely to be the target of CRD IV and CRR, they will find themselves less affected than smaller proprietary trading firms.
"Bad things will happen"
So let’s assume that a firm needs to calculate its capital adequacy requirements; how does it do so?
Well, it can build its own calculator (some firms have) or it can ‘buy’ one from a vendor. Beware, however, that one vendor’s application of the model to a calculator might differ from another’s, giving (perhaps wildly) different answers.
If the FCA does not agree with a firm’s calculations, bad things will happen.
Is it any defence for a firm to say that it used a trusted and well-known vendor’s calculator? Maybe but probably not - it’s a lack of control over the quality of external providers.
So should a firm build its own calculator? That would depend on costs, I suppose - because if a firm’s own calculator ‘gets it wrong’ it’s the firm which is firmly on the hook for that.
So where next? Let’s hope that the FCA looks at the different models out there and the ways in which the industry reduces the probability of a firm blowing up and/or a firm blowing up and taking the market with it.
There may be some advantages in firms that will need to become authorised and regulated under MiFID2 waiting-and-seeing. However, the smart (maybe not ‘smart’ - the ‘cynical’) money is on the FCA not agreeing to a firm applying an internal risk model without it having run standard models for some time first or having a track-record of applying successfully an internal risk model.
So for firms that are contemplating becoming authorised and regulated under MiFID2, perhaps the best thing to do is to set up asap and run an unregulated business for a period of time to gather data to show the FCA (when an application is presented to it) that show the firm’s own risk model ‘works’.