Quantum Computing2 Comments

Excerpt from an interview that took place at the Annual Global Derivatives Conference, May 11th 2016, in Budapest (Hungary).

1.     What is a quantum computer?

A quantum computer is a computer that exploits the basic laws of quantum mechanics to solve mathematical problems. Nature solves extremely complex mathematical problems all the time. Consider, for instance, the simultaneous computation of the trajectory of planets in the solar system. This is an extremely complex mathematical problem, and yet Nature solves it for free and instantaneously. How? Digital computers store information in bits, which can take a value of either 0 or 1. In contrast, quantum computers store information in qubits. Qubits are in a superposition of the two states. This means that they store a 0 with probability p and 1 with probability 1-p. Because of this quantum property, they can explore an exponentially larger number of combinations compared to digital computers.

2.     How does quantum computing give you an edge?

In two ways: First, quantum computers will allow you to solve problems that digital computers could not crack in millions of years. Time is money, and exponential speed-up gives you an unparalleled edge. Second, quantum computers are non-deterministic, probabilistic devices. Think how banks assess risk every day. They use digital computers, which are deterministic devices. Using a deterministic device to simulate risk makes little sense. Instead, we should use non-deterministic devices to operate directly (rather than simulate) random variables. That’s what quantum computers will do for finance.

3.     What are specific practical applications in finance?

There are at least four immediate applications:

·        Generalized multi-horizon portfolio optimization: This is a NP-complete problem, which means that it is intractable to digital supercomputers.

·        Clustering: Current clustering methods apply heuristics, which lead to suboptimal solutions. Powerful clustering methods would allow regression analysis to apply modern discrete mathematics, rather than linear algebra or calculus.

·        Scenario simulations: A quantum computer could generate an extremely large number of scenarios and evaluate the distribution of outcomes.

·        Option pricing: Path dependent derivatives can be priced using quantum computer’s power to evaluate a large number of combinations.

4.     How do you use quantum computers in your work?

I cannot detail how I’m planning to use them in our investment processes. But let me give you a general idea. Markets are complex probabilistic networks that require tools and techniques adequate to capture such complexity. Quantum computers are probabilistic devices that solve problems encoded as graphs. It is a great match. Moreover, in today’s world, if you want to have an edge, you need to search for alpha outside anybody’s reach. You need to be able to solve problems that nobody else can solve. Digitalquants will be at an insurmountable disadvantage, when they are forced to compete with the coming generation of quantum-quants.

5.     Who are the main players in this area?

D-Wave is the first company to commercialize viable quantum computers. 1QBit is the leading quantum software development company. Both are many years ahead of their competitors. 

6.     Is this technology ready for use today in a business setting, or is it still in an experimental phase?

Probably not yet ready, but very soon. Of course, once quantum computers are deployed in production, it will be impossible to catch up. Alpha will dry up overnight, and most traditional hedge funds will go bust, as the old inefficiencies disappear. In particular, many high-frequency traders may starve, as large asset managers will produce robust portfolios will little noise-trading.

7.     How can our viewers become more familiar with this technology?

Very easy: Join the www.QuantumForQuants.org community. Don’t get left behind.

2 Comments on this article

  • Michal May 12, 2016

    Oh yes, thanks for your kind care on denoising this great source of further Q4Q knowledge. Both of my posts ( + this one naturally too ) are of no further value in this direction, the second one was posted before ‘ve noticed your re-confirmation of the observed fault I did on viewing just pre-cached pages. Once more thank you Marcos.

    n.b.:
    I will come with more details on the intended subject later this week ( as you’ve added the limited DoM / L2 OrderBook mechanics are clear and phantom-(il)-liquidity faked by (some) HFT prop shops are making this subject even more cardinal for ( let me call it ) a “dumb”-force rebalancing attemps ) but the core reasoning of my doubts on achievable AUM performance-envelope / portfolio rebalancing practices is by far not abstracted from realistic transaction costs and observed composite slippage effect of both natural ( order-matching consumed )+ phantom ( JIT-cancelled ) liquidity.

    Anyway, this time, it will be properly directed to the original place.

    Reply

  • visit this website January 15, 2017

    Great post. I am facing a couple of these problems.

    Reply

Add a comment