CI Futures in Bloomberg Terminal - YouTube

Hi everyone. This is Tony Nash with Complete  Intelligence. I'd like to walk through our app  

CI Futures that is on the Bloomberg App Portal.  Complete Intelligence is based in Houston, Texas  

and we've built a completely automated, globally  integrated artificial intelligence platform  

that looks at cost and revenue forecasting. We  look at markets and what we're looking at is  

better ways to help people understand future  prices and costs. One of the key issues we've seen  

with Covid and with all the volatility and  uncertainty in the market is companies and  

investors trying to understand how they plan for  that environment. Over the past year, we've had  

a number of companies come to work with us to  look at their revenue planning, to look at their  

exposure to commodities, and to look at their  overall exposure to markets. We've used CI Futures  

which is available on the Bloomberg App Portal  for example with a global chemicals firm. And  

with that we've helped them reduce their forecast  error rate for one of their major products to 4.4  

percent on an absolute percentage error basis. At  the same time, we've helped a global mining firm  

reduce their error rates on their gold  forecasts by 38 percent. As you can see,  

those two companies have done extraordinarily well  in equity markets this year. Whether it's directly  

a result of that, we’re not exactly sure. But the  message here is that companies that take analytics  

seriously, who deploy say artificial intelligence  solutions can respond better to markets and to  

customers. As we've put together CI Futures and  as Complete Intelligence has grown as a company,  

we've really built the company on the basis that  the status quo planning process and status quo  

forecasting process is just flawed. Many  companies and many investors rely on  

consensus forecasts which are very similar to  forward curves. And consensus forecasts have a  

fairly high error rate, 20 percent, in some  cases more, in some cases slightly less.  

But generally, consensus forecasts have  a high error rate. People take those in  

when they're planning their portfolios or  planning their investments or their budgets  

as a benchmark for their activity. What we've  seen is with Complete Intelligence, we have on  

average of 4.6 absolute percent error rate with  our forecasts across currencies, commodities,  

and equity indices. Those that are available on  the Bloomberg App Portal. We've helped companies  

really plan and save millions of dollars and we  help investors understand their risk exposure.  

When we look at consensus data, a real tangible  comparison we do is look at the full year  

2019 consensus forecast to our Complete  Intelligence forecasts. These are actual  

consensus forecasts taken out of the report  referenced below this table. And we look at  

the consensus forecasts in the yellow column. The  Complete Intelligence forecasts in the blue and  

actual forecasts in the gray. When we look at  the right-hand side of this table, we see that  

Complete Intelligence beats consensus forecasts  88 percent of the time. There's a tangible metric  

there to show how we've performed across say  crude oil, industrial metals, and precious metals.

When we do this, Complete Intelligence, CI Futures  is really powered on publicly available data.  

We do two things. The first thing we do  is take in publicly available data. It's  

about 15 and a half billion data items  and we've built a very large ecosystem  

to understand how the world economy and how global  markets work. We run that process twice a month  

and we're taking billions of calculations to  understand price action. We work on what's  

called an Ensemble Approach. We take fundamental  methodologies. We have technical methodologies  

as well and we look at the best configuration of  those methodologies to forecast the assets that  

are within CI Futures. It's also possible to take  in custom assets from investors and from companies  

so that we can help them understand their  specific items that they're looking to forecast,  

whether it's a cost, whether it's an investment  asset, whether it's a sales revenue item,  

something like that. We can take those  in from customers and we can put them  

on a closed login for customers so that they can  understand the forecast outlook for those assets.  

With our methodology, as I said we're  using an Ensemble Approach with billions of  

items within the context. We actually utilize  about 10 thousand scenarios for every asset we  

forecast. So twice a month, on the first  of the month and the 15th of the month,  

we're gathering all the publicly available  data that we use. We're testing every line item  

every month. We're weighting that line  item and reconfiguring the algorithm and  

then we're reforecasting those assets.  With every asset, the market movements,  

the movements in the world economy, the movements  and things like world trade and so on and so forth  

are taken into account so that we can understand  what that cost or what that asset price will be  

every month for the next 12 months. So  CI Futures is a subscription service on  

the Bloomberg App Portal. We have a very simple  interface. What we want to do is present users  

with a monthly interval outlook over the next 12  months. We provide high, base, and low scenarios  

within that forecast context. Now, the base  and low scenarios are one standard deviation  

applied equally over 12 months. We're not running  discreet scenarios for a high and a low scenario.  

We're taking the baseline, the likely scenario and  and putting a one standard deviation band around  

it. What's really key for us here is that 97 of  our forecasts typically have less than 10 percent  

absolute error. 84% have less than  5% absolute error. We really focus on  

error performance and with every iteration of  our forecast, we're collecting those error data.  

We're understanding how we perform to market and  where necessary, we're making system-wide changes  

to take account for say volatility or say  government intervention and so on and so forth.  

What's key to understand here is we are  not manually adjusting any individual asset  

within CI Futures. We continue to iterate the  overall ecosystem of CI Futures so that we can  

best account for volatility. For example after  Covid, we introduced much more volatility-specific  

methodologies so that we could identify ways  the volatility would impact market prices.  

With CI Futures, we're really trying to help  companies and investors do more with less. We have  

a subscription available for individuals on the  Bloomberg App Portal. We're trying to help people  

improve net income, improve their cash flow,  and improve their valuation or market cap. It's  

really important for us to help our clients and to  continue to understand how these assets move. If  

you can check us out on the Bloomberg App Portal,  we'd really appreciate it. Thanks very much.