To capture the market-timing, pricing inefficiencies and structural relationships targeted by most multi-asset portfolios, investment managers build a strategy using a carefully constructed set of products within and across asset classes. This set of products is measured and executed as a trade “bundle”. Each product in this bundle has a deliberate impact on the fund’s ability to obtain the intended investment objective and the strategy itself is not realized until each leg of the bundle is fully executed and bedded into the portfolio.
Multi-asset portfolio managers looking to optimize their market impact and cash utilization must carefully craft the composition of and sequence the order by which they execute the legs of their trade bundles. Inputs that feed this composition and sequencing are pulled from all aspects of the trade lifecycle. This differs dramatically from long-only and single asset strategies that typically rely only on pre-trade and market-timing information. To facilitate this more complex and multi-variable trading approach, multi-asset strategy firms have tended to invest in data modeling and distribution tools that provide them an edge over their more product-aligned service partners.
The foundation for these buy-side tools are proprietary robust data models that allow investment managers to easily add new asset classes and product types. This process is usually managed by the middle office and on the whole, the relationship between the front and middle office is much tighter in multi-asset than in single asset trading firms.
Proprietary pricing, valuation, risk, credit and collateral models sit on top of the robust data foundation. Information from these models is blended with real-time lifecycle outputs and delivered throughout market leading organizations via real-time dashboards. This allows the buy-side to work from a more comprehensive information set than their sell-side partners and leaves the sell-side poorly positioned to rebut challenges as their information set is still cut by product silo and delayed due to batch cycles and incompatible core data.
