Using investment analysis and technology to get (and stay) ahead of the curve
Bloomberg’s Daybreak podcast is a great device for anyone market focused and looking for multi-daily updates on what’s happening across the world. A topic that that’s been on the agenda of many an investors mindset over the last few months, and was discussed in today’s podcasts, was the effect (or should I say lack of effect) that tariff talks have had on trade sentiment and the markets in general. Cameron Crise of Bloomberg discussed this in today’s episode of his Macro Man podcast (available on SoundCloud HERE) and how there were frequent questions as to why the US markets don’t sell off despite the tariff threats. He goes into 2 reasons why and it’s the second of these that I find most interesting, as it relates to work I had been doing at QMG and macroeconomic analysis.
His first reason was quite simply that there is no longer any advantage of buying any dips in prices across the futures markets because everyone seems to be doing that now and anytime there is a dip, even a minor one, will see a deluge of buyers. Basically, the opportunity here for that has dried up.
The second, and more pertinent point as to why the US markets remain so toppy, is that corporate earnings remain robust. This is somewhat perplexing considering that imposed tariffs should be hitting corporate bottom lines. However, earnings and expected earnings continue to rise. There was also mention of the limitations of sell-side research in predicting future earnings however that this does move the dial quite significantly when there are earnings upgrades or downgrades. The reason that a limited predictor is given such strong weight is that investors have little else but these earnings predictions to act as their investment guide. In the US, corporate earnings occur every quarter which means that there is a potential 3-month gap between updates on stocks you follow (barring any intra-earnings season events). Most investors are caught up in this spiral of having to rely on this cycle but at QMG we showed a way out.
The QMG data and process should be of interest here because it provided a potential solution for investors stuck in the doldrums of awaiting the next piece of research or next corporate earnings announcement. QMG data was released monthly (see QMG data model overview below) and with the US being the flagship marketplace, provided a peek into corporate earnings months in advance of company announcements. The dataset gave investors a guide as to the direction of corporate earnings at as low a level as we could glean from the US statistics agencies (from whom we would get the data). Because this data was provided monthly, it meant we were ahead of the curve and flagging potential inflections (or supporting bull/bear arguments) in various sectors. This is all well and good to get an early indicator, but it would be nothing if it were not useful and we had to test this somehow.
The key piece of evidence that showed me and my other colleagues that this worked came down to statistics. Knowing very little about the markets 4 years ago (other than what Hollywood showcased or what documentaries delved into), I not only had to implement a new analytics infrastructure at QMG, but I also had to play a lot of catch up my market education. Statistics was something I exceled in during my university days so it was nice to see how important a solid understanding of this would be for our analysis. Coupling this with my knowledge of automation software made for a perfect match. In the case of QMG, we were shown on various Excel files how closely correlated related some sector mixes were when paired with actual company earnings. This made sense considering that the companies that investors and analysts follow are also those same companies that have to report their activities to various government agencies, who in turn create the national statistics that are the key ingredient for the QMG dataset. The correlations were good evidence to showcase that the data was a leading indicator of market direction for US stocks but the analysis in only an Excel format meant the spread of this news could only be limited. This is where QlikView stepped in.
QlikView was a tool that I had used in my previous work as a data consultant and whilst it is seen as mainly a reporting and dashboard design tool, perhaps the biggest feature is in how it allows for automation across repeatable data processes. If we had kept on using Excel to showcase the close correlations between our datasets and corporate earnings, our stock and market analysis would have been limited to reviewing companies one by one. Using a tool like QlikView, I was able to couple its statistical functions with its ability to automate and aggregate these calculations and piece it all together in a graphical interface. Using various filters and charts, analysts (internal and external) could quickly see which correlation relationships were the strongest and how these changed over time. Below is an example of the QMG data (called INSIGHT) for a number of US sectors that relate to the machinery producer Caterpillar and how closely that relationship was for sales growth over nearly 20 years.
There were some great calls we made during our time that the rest of the investment community had likely not seen (as evidenced by the contrast between our calls and the average of analyst earnings estimates) and this was only possible thanks to the combination of a unique approach to market analysis (the QMG process) sitting on top of state-of-the-art aggregation technology (QlikView). There were other software products that could have done a similar job for us at the time but for the price and features available (as well as my background of using this tool) we found QlikView to be the right thing for us. They’ve since made further enhancements with their Qlik Sense product which we also used but the point is that technology continues to push the capabilities of investment analysis forward and more people and organisations need to take advantage of this.
Going back to the investment analysis and tariffs, the key element to watch, according to Cameron, is US earnings and their future trends. The QMG dataset offered a good way to watch this and perhaps now that equity research is becoming more data and technology focused, it will make a resurgence. I’d certainly like to see that early warning indicator. Wouldn’t you?