Serving the Quantitative Finance Community

 
User avatar
CommodityQuant
Topic Author
Posts: 57
Joined: July 5th, 2007, 6:16 am

Possible problem with Quant Dev consultancy business model

April 30th, 2017, 2:25 pm

Consider a scenario where a bank (call it AnyBank) hires a consultancy (call it AnyConsultancy) to do some quant dev work.
An AnyBank quant (call her M) therefore passes on instructions to the quant dev consultants from AnyConsultancy.
Obviously, the project's success or failure hinges on the quality of the cooperation between the AnyBank quant and the quant
developers from AnyConsultancy.
The incentives of the developers from AnyConsultancy are clear.  They want the project to go well which will lead to strong
feedback from AnyBank which will be good for their careers.

But what is the incentive for the AnyBank quant, M, to give clear instructions?  If the project goes badly, M will
blame the consultants, the AnyBank managers will believe M, and the AnyBank quant can then make 
a persuasive case that the way forward for the project is for AnyBank to hire quants, rather than rely on consultants,
and M is then likely to be a manager of a group of new quant employees and M will therefore gain in seniority.
M is therefore incentivised to give poor instructions so that the project fails, and so that she becomes a direct supervisor
of new AnyBank quant employees under her management.
In practice, M won't deliberately sabotage the project by giving obviously wrong instructions.  M could simply refuse to 
spend any time thinking about what she says, and never provide any written documentation.  Such sloppiness will
lead to the project failing which will be good for M as she then argues (as above) for her to directly supervise 
AnyBank's own quant devs which will be better for her career.

What can be done, and what is commonly done in practice to prevent the AnyBank project failing in this manner?

Many thanks for your responses.
 
User avatar
snufkin
Posts: 64
Joined: January 25th, 2017, 9:05 am
Location: Cambridge

Re: Possible problem with Quant Dev consultancy business model

April 30th, 2017, 3:32 pm

Well, the number one rule of consultants, be it a big consultancy or small is that if the project fails, the consultant takes the blame (but not the bill). Given that all those consultancies are still around, it seems that the problem is not as big as it sounds.

But your description does indeed sound scary: the success of such an important project hinges on one naughty quant! And he or she is such an asshole that it seems that he or she is willing to wait for months just to make the consultancy fail!

Something isn't quite right here, is it? First of all, how come she even talks to that poor consultancy in the first place? Who let the consultants in the bank? Why has it happened? 

So the first answer is, know your counterparts. You need to know people all around the organisation; you need to be friends with all the PAs, quants, devs, managers, their managers, etc. — because you will need all the help you can get. 

Second, making project fail is a huge risk. If you have a consultancy working on a project that you're responsible for, you normally are interested in project's success; it does not really matter whether you can shift (part of the blame) later on. So of course M's interested in getting the project done. As for the career, I don't think getting more directs is a wise career move: it's much more headache, and hardly any benefit compared to managing a contractor. In my experience, the best situation you can get in is being an MD without any directs.

So the advice would be, assume best intentions. Most probably, the goals are already aligned. If it's not the case, find the sources of concern — don't ever guess, ask, ask, ask, and ask again — and see whether the concerns have any ground. If M is genuinely malicious (which, as I said, is very unlikely), escalate that — M is not the only person in the firm, and M wouldn't let you in in the first place. If that doesn't help, cut the losses and run away.

Hope that helps.
 
User avatar
DominicConnor
Posts: 41
Joined: July 14th, 2002, 3:00 am

Re: Possible problem with Quant Dev consultancy business model

May 18th, 2017, 3:22 pm

The number of possiblities for what is going on is large,  so I'm only going to sample.

Firstly, I'd ask why they hired someone in ?
If they were happy with what the quants were doing, why spend the money ?

Is that they aren't delivering, are too expensive or the volume of work has gone up ?
It sometimes is the case that a consultant is hired simply because there is something they don't know how to do and haven't the time to learn.

Their incentives are complex and unknown, but I'm usually sceptical of the "let others fail" hypothesis for prediciting behaviour. If this goes wrong, shit will fly about and stick in random places, both fairly and unfairly. That being said, I've seen it happen, so I'm with snufkin on this.

A good soft skill technique is thanking people.
When you get actual help from the quants, drop a line to them and their manager thanking them. Nothing gushing,. just a quick note, not too often.

That will build you good will and create the impression of you being a good team player.
It also of course allows for greater contrast if they start throwing shit around.

Also of course if they've told you to do something wrong or useless it draws attention to it.

Conversely do not send whiny emails copied to managers. Ever. This destroys trust and makes enemies.

One technique I've shared before on Wilmott.com but not recently enough is "celebrations as a weapon". Define some milestone, doesn't really matter what as long as it is credible. When you reach it, bring in donuts, chocolates or buy beers. Again there is a frequency optimisation based on local conditions, thank those that have helped you. You become associated with delivery of success (even if no one really understands what you do), and with goodies.