What's the next big thing in trading?
In my estimation, it's The Cloud.
In my estimation, it's The Cloud.
The Cloud will have as great, or even greater affect on trading as the Internet.
That's a big statement, and I bet that many of you, are nodding your head in agreement.
The Cloud is going to change the way that we think about Intellectual Property, whether we need to own it, buy it, or simply lease it - which is the option that I think will emerge as the de facto standard in 50 or so months.
In the blink of an eye The Cloud will exponentially increase the amount of data that we back-test against and thus, will open up true data mining to the masses.
No longer will quant shops have exclusive access to ripping through the massive amounts of data that really drives the markets.
Creating tool sets and algorithms for these data sets is going to be a great gig for a the tech folks who hustle and early on stake out their corner of The Cloud.
New brands that are cloud-centric will spring up and will provide their owners with a marketing edge for at least a decade or so.
Here's a first pass at what 'before' and 'after' The Cloud will do for Joe Trader:
Pre-Cloud Retail Trading Architecture:
Post-Cloud Retail Trading Architecture:
Brokers and software developers making Cloud services available to their clients will wipe the deck with old-line brokers and trading platform developers who don't adopt to the new model quickly enough.
We are working towards a product that is a layup to Cloud services. We will be offering automated trade management for news sentiment algos and leasing of algo IP via slew of .NET web services and will be posting more about it here soon.
Anyone know of any shops offering Cloud Services yet?
Hi Carl
Good post. Its exactly this view of the future that led me to set up a meetup group in London 9 months ago (London Cloud Computing Meetup -> http://gridcomputing.meetup.com/4 ).
Here are a couple of non-functional requirements which you will need to consider.
Reliability
Latency
Security
Scalability
The quality level of each of these will be based on the SLA you are prepared to negotiate and pay for with your cloud infrastructure provider.
We talk about these issues at our regular meetups.
I've been developing application infrastructure at my current client which encompasses the "Market Data from All Exchanges", has a lot of elements of the "Algorithmic Database" - and also provides an algorithmic workflow as described in your diagram above. Its based on an internal cloud which can be easily deployed externally - or migrated to a hybrid internal/external setup.
Cheers
Paul
Hi Paul,
I concur with your points that no matter what service you are offering and know matter how clever your approach is, you better have the day-to-day stuff that makes software viable in place.
Very cool that you are developing an internal cloud that can be ported externally if there is a reason for it...which I am sure there are as The Cloud does not yet seem to have any type of Trading App critical mass yet.
All the best,
Carl
Hi Paul,
I concur with your points that no matter what service you are offering and know matter how clever your approach is, you better have the day-to-day stuff that makes software viable in place.
Very cool that you are developing an internal cloud that can be ported externally if there is a reason for it...which I am sure there are as The Cloud does not yet seem to have any type of Trading App critical mass yet.
All the best,
Carl
Carl-
Well thought-out, and I do believe that a move to this type of paradigm will likely occur. In the meantime, however, we will be dealing with the current structure which - having been in place for some time - has become a cash cow to the participants. All the communication protocols, pricing structures, and interpositioning of firms who have the capability to hoard information will serve as a roadblock to innovation in this sense.
To add to what one author wrote above, it will not only need to have the "day-to-day" stuff in place that people use, but it will have to be done in a way that first can either price out current providers (or entice them to change), and price themselves IN to a market that is highly dependent on the current way of business.
What is (1) the total cost, (2) the benefit, and (3) what can go wrong?
1. What's the total cost? In learning curves, development, optimization, firewalls and domains, management, QA, risk, software, tools, infrastructure, operations, ...
People were recenty trumpeting SOA as the next great thing, and then slowly it began to sink in that it costs a multiple of simpler approaches, and the benefits are slim unless you really are assured of a lot of reuse. Which brings us to ...
2. What's the benefit? Who really wants "unlimited processing power" and "unlimited data" or whatever it has to offer? Not a lot of uses. P.S. I've been a quant strategy developer, and I found quantity of data to be processed is only sometimes an important bottleneck.
When people first see a hammer they often seem to think everything is a nail. Remember client/server and C++ and neural networks and all previous wonder technologies (including SOA). At first everyone thought they were solutions for everything, and then in each case it emerged that each tool is only good for a specific set of problems. Clayton Christensen talks about the three stages of a new technology, the final stage being a tested theory of when the tool is actually useful or not.
3. What can go wrong?
How do you find a bug in a cloud? What if one buggy or leaky component impacts all the others? How do you manage versions? What management and coordination do you need to make it work? What unique QA problems does this model create? What other unique risks do you have to consider? How many computing-intensive applications are shared by enough people to bother?
TH: Good points, all. As to your last question, I'm not sure that the sheer computing power of the cloud is the most important part, nor the shared applications. As a retail trader, the part that would interest me most is easy access to a large, clean, up-to-date source of market data, specifically for index and equity options.
Keeping end-of-day data for equities and their option chains is not so bad; but minute-by-minute, let alone tick-by-tick data, uses vast amounts of storage. Real-time updates to option prices require a very large (and increasing) amount of bandwidth. The cost of the data can be very high. And writing market data adapter programs that can deal with errors, missing data, stock splits, symbol changes, new instruments (such as binary and weekly options), etc. is enormously complicated and time-consuming.
I'm not interested in doing anything that's terribly computing-intensive, other than calculating theoretical values and option Greeks, and some basic query processing. But I think a service that could provide relatively inexpensive access to a shared pricing database would be of great value.