I read numerous postings on this forum about these different packages. We are investigating the possibility of buyng vs building a derivatives valuation system and I would like to get feedback.Principally we are interested in using the system for valuing equity exotics and structured products, so it is primarily this ability we are interested in, we have our own model for vanillas which has served us well. That being said:1. Does anyone have opinions on which of the systems might be the most worthwhile for this application?2. Is the cost of these systems (quite high!) really worth it, or can one hire developers and do it better and/or cheaper?3. How difficult are the systems to actually implement and being using? Are we talking a 1 month project? 3 months? 6 months? How many IT personnel are required for support/maintenance?Your thoughts are much appreciated.Rgds

Hi,I know its been a while, i just want to find out if you got any response to your questions as i am also trying to compare both calypso and summit in terms of its stp functionality in the derivative world

I have played around with Murex for Equity Derivatives but not within the MACS framework (yet!). How Exotic were you thinking?The TradePad is good for building Structured Trades based on existing products and there is always their API if you want to implement your own models.

I've been in the business a while, worn many hats and my experience tells me that while one tends to get better hard analytics with internallly made software, just about everything else (risk, P&L, reporting etc) is grossly inferior. Vendors tend to be much better at generic software attributes like data management, extensibility, reporting etc. but comparatively behind the curve on analytics. My advice is to not reinvent the wheel and look for a vendor system with a well developed API that you use to integrate your proprietary analytics. All of the vendor systems you've mentioned are serious stuff and I think the Forum colleagues would need more understanding of the markets you work in to offer you much more insight. An install of one of those things would be at least 3 months in my opinion but more like 6.One thing you should insist on when doing due diligence on prospective vendors is that they give you contact information not only for firms who successfully use their software, but also for firms that stopped using their software (there are always some!). Learning why an installation failed can be quite enlightening and yields many insights into post-sales treatment of clients.

All are :OldExpensiveDifficult to implementIf you have people who sucessfully implemented one of these product, then go for it. Else, you will pay K€ / even M€ to get simple things working. (licensing cost and consulting, and implementation and ...)The fact is that these product are so expensive and heavy that you won't find any user telling you they are not happy.When you've put 10 M€ in a product to get vanilla product booking and valuation working, you go low profile.In 2009, you'd better go agile and build up a team of highly reactive and experienced developpers. Especially if you have a precise idea of what you need and what you want to do. It will save your money and your time.

Murex: (Top of the pops - In the Vendor world) - has its limitations - I just posted some of them in another thread today).Summit: Summit MUST - Is apparently pretty good for exotics. BUT - It only has one customer that I'm aware of - KFW in Germany. They were driving a lot of development in the past.Sophis: If you have equity-only then this might be a way to go. Habe not heard much about them recently.Calypso: They have a poor structuring tool eXSP that is arcane and difficult to integrate to. Nice simple API's for model integration albeit they screwed up the pricer/model relationship.Don't agree with fredba that these are all "old" - not necessarily the case. But - they are expensive and time consuming to integrate. If all you're after is a valuation/risk platform there should be other more lightweight tools out there that you can look to. These systems are overkill...SuperDerivatives have the asp route sewn up for FX Vanillas and some of the structured deals. Not so sure about eq. That said you've no idea what model they're using to value or how they calibrate the deals but anyway.Alternatively - NumeriX and Pricing Partners have various tools. XLLs, SDKs and lightweight apps that allow the user to perform valuations on arbitrarily complex deals. Pricing Partners have recently opened their code up to customers and NumeriX, I believe, have a model API that users can plug-in to.

When I mean old; I'm not talking about the way they have been revamped using gui technology.They are old product, they all experience troubles with their client on the same subjects :- scalability- releases- maintenance- qualitySome mainstream IB have dedicated team of 100+ developpers on these products.The business model is more like : on the day you sign the contract, you start an implementation that will never end. the fact is that in a 6 month window, your business, the market, everything may evolve, so you will always be running a project portfolio.Don't misunderstand me : these products are good products made by brilliant people, but, they are also selling their product and support/Impl/Prof services.for the vendor, something that should be your concern is :- Really used technology behind the software- If the company is listed (that's very important, being listed as a software provider in this business is generally a bad idea, they are working on quarterly basis, and will cut-down r&d investment)- r&d will always be at least 15% of revenues. Get the details. 15% to 20% is a "standardised figure" that every provider is using. But if the accounting figures are true, it is not always true that they are doing real r&d stuffs- number of specific versions (are you going to be a new one)- product plan over 4 years (yep, not the next quarter sales opportunities)- last implementation (start date ,end date, contact to check)- 5 biggest clients and their contribution to the revenue

scalability, releases, maintenance & quality are all software issues independent of whether its a vendor solution or an in-house developed system. Over time the scalability will be questionable and the maintenance and quality certainly will be. I've seen one instance where the bank didn't actually have the source code of the , in-house developed, system that was running their fixed income books. Scary stuff.I agree with all your points otherwise - but have you really managed to get a 4yr roadmap out of a vendor? Sounds like a vapor-map!?!?

Don't forget the etc.

No I didn't. Exception made if they ere 're-ingeneering' the product.It is just a good way to figure out if they are reliable.If the guy answer "No", he is honnest and reliable"Yes, you'll get it by the end of this week", you're talking to the sales guy"Yes, it is in the presentation. It is part of our strategy.", I don't know who you are talking with, but run, don't walk, away.The most important things are not the question, but the reliability of the answers. (very much like a DDQ indeed)QuoteOver time the scalability will be questionable and the maintenance and quality certainly will be --> yes, and that's exactly why the true age of the product is a meaningfull information.

Hunting, have a look at our font-office and risk-management system PLaNETa.Unfortunately, information on our English-version website is a bit out-of-date. At the current time, PLaNETa supports spot and forward trades, vanilla and exotic options on equity (asian, barrier, lookback), binary options, and structured products - we have a structured products designer - the user enters the payoff function of the product, and the software automatically "reverse engineers" the product, lists the vanilla and binary options it consists of and calculates the value of the product.Risk management features include Value-at-Risk, What-if analysis, etc.You can see video of the software - the comments are, unfortunately, in Russian, but the GUI is in English - so I think you will get a general understanding of what the software does (though those videos cover only a small part of the features of PLaNETa).Value-at-Risk calculation in PLaNETa (video presentation)Constructing a structured product in PLaNETa (video presentation)The important thing is that PLaNETa can be provided with full source code. We also have an Excel-like Add-Ins model which allows clients to write additional modules to PLaNETa to extend the software. This is flexibility which you certainly cannot find in software offered by companies like Murex. And the price of our software is far smaller than that of systems like Murex, Sophis, etc.

QuoteOver time the scalability will be questionable and the maintenance and quality certainly will be --> yes, and that's exactly why the true age of the product is a meaningfull information.---> Are you saying that you'd take a product in it's infancy?? In my experience the architecture if frequently in reasonably good shape early on - but the financial layer is less so.

No, but some products exists because they were internal development.It is just a question of lifecycle :- If too early, you may be more a sponsor- If too late, you may face problement of maintenance, scalability and reactivityAnd yes, the real "issue date" of a product is a good indicators to keep in mind.It depends also what kind of client you want to be :- Big client with bug influence on product strategy- 'trend follower'- Full service or on site with internal IT..I think that in 2009, no one is expecting a pricing library out of the box from a software provider. Everyone is expecting simple and evolutive product, with interoperability functionnalities. And it is an art to keep a software 'lightweight' but yet very flexible.

In the last few years all the vendors have been aware that few institutions want to spend much money on the implementation. They have started to provide ASP (Application Service Provider) to their offering where they provide a pre-configured database and web access to use their applications. This makes it somewhat easier for smaller institutions to use the software.