Alternative facts. And data.
Translation of the "Portfolio focus" column in Kapital, Norway's largest and leading business magazine, June 14th 2019 edition.
There is an insane amount of data out there. Zettabyte (ZB) is a measure of digital information (zetta is 1000 to the 7th power). It has been estimated that all words ever communicated by mankind could be stored 200 times within one ZB. 10 years ago, the entire datasphere - defined as all information on the internet – was about half a ZB. Today, when most would accept that this internet thingy has come to stay, the datasphere makes up about 50 ZB. This is expected to grow beyond 100 ZB within 3-4 years.
There is hardly a Nobel prize incoming for concluding that there is money to be made for those able to sort out data gold from cute kittens and the various bits and pieces of the Kardashian family. For institutional investors, the ability to manage massive amounts of data has become the main activity. Leading asset managers increasingly resemble tech and data companies. The Swiss company Sentifi has estimated that only 10% of investment relevant data is to be found through traditional media (Bloomberg, Reuters, the financial press, etc). 90% is to be found through alternative sources, and that does not even include president Trump’s alternative facts. This would include data collected through social media, from satellites and drones, the internet of things (IoT), and so on. There are currently about 5000 satellites in orbit around our planet, many soon capable of reading the newspaper you hold in your hands, there are 7 million drones capable of seeing and measuring pretty much everything, and there will soon be 10 billion IoT appliances that can talk to each other as well as to your jeans and your nifty Nike trainers. Just think about that the next time you’re alone, but feel as if somebody is staring at you – it could be the shiny new Miele fridge taking a peek.
Information equals power. There is a huge difference between asset managers that can actually use alternative information, and those who just shake their heads and fold out the FT. Let’s take as an example the soothsayer business, or the macro economists as they often call themselves. In the old days, i.e. 2 years ago, the essence of their spiel was to follow a number of key macro data; growth, inflation, balances of trade. Based on these key data, a prognosis was made regarding the future development of the world. In turn these prognoses could drive markets. Today, however, it is increasingly the case that «own» data is used to estimate in advance how the official data will develop, and when the official figures do appear, positions have already been taken once or thrice. As examples, and as discussed previously in these pages, one may estimate consumer activities through real time drone data covering the number of cars parked outside shopping malls. Trade activities may be estimated by using real time and global satellite surveillance of which ships are and go where. Consumer satisfaction with Apple may be measured by AI and analysis of social media comments. And the list just goes on. The concept of “forecasting” is being replaced by the idea of “nowcasting”. The only remaining question is really whether an asset manager has access to, and the capability to manage, really big data.
Investors chase investment returns. Returns may come from beta, which is what the market gives you for free, and alpha that comes from beating the market. Beta is available from cheap or free (yeah, right…) index funds, and alpha can be bought from the best and the more expensive active managers. A special, but large group, are those who insist on buying what is really beta from expensive managers that were supposed to deliver alpha, but that is just silly. However, today even alpha is just not alpha. Using AI, learning models and big data processing, it is perfectly possible to identify systematic sources of alpha that can be easily replicated, such as long / short, carry, value and momentum strategies. These «false» alpha sources should not be too expensive to access, and there is an expanding universe of managers and funds providing systematic alpha at sensible prices. That leaves a pretty short list of managers that actually both charge for and provide real alpha. And the rest, who neither deliver beta at beta prices nor alpha at any price, should probably try fairly smartly to swap some of the soothsayers for members of the hoodies, Converse and portables-covered-with-weird-stickers legion. While they still have some assets left to manage.
Translation of the "Portfolio focus" column in Kapital, Norway's largest and leading business magazine, June 14th 2019 edition.
There is an insane amount of data out there. Zettabyte (ZB) is a measure of digital information (zetta is 1000 to the 7th power). It has been estimated that all words ever communicated by mankind could be stored 200 times within one ZB. 10 years ago, the entire datasphere - defined as all information on the internet – was about half a ZB. Today, when most would accept that this internet thingy has come to stay, the datasphere makes up about 50 ZB. This is expected to grow beyond 100 ZB within 3-4 years.
There is hardly a Nobel prize incoming for concluding that there is money to be made for those able to sort out data gold from cute kittens and the various bits and pieces of the Kardashian family. For institutional investors, the ability to manage massive amounts of data has become the main activity. Leading asset managers increasingly resemble tech and data companies. The Swiss company Sentifi has estimated that only 10% of investment relevant data is to be found through traditional media (Bloomberg, Reuters, the financial press, etc). 90% is to be found through alternative sources, and that does not even include president Trump’s alternative facts. This would include data collected through social media, from satellites and drones, the internet of things (IoT), and so on. There are currently about 5000 satellites in orbit around our planet, many soon capable of reading the newspaper you hold in your hands, there are 7 million drones capable of seeing and measuring pretty much everything, and there will soon be 10 billion IoT appliances that can talk to each other as well as to your jeans and your nifty Nike trainers. Just think about that the next time you’re alone, but feel as if somebody is staring at you – it could be the shiny new Miele fridge taking a peek.
Information equals power. There is a huge difference between asset managers that can actually use alternative information, and those who just shake their heads and fold out the FT. Let’s take as an example the soothsayer business, or the macro economists as they often call themselves. In the old days, i.e. 2 years ago, the essence of their spiel was to follow a number of key macro data; growth, inflation, balances of trade. Based on these key data, a prognosis was made regarding the future development of the world. In turn these prognoses could drive markets. Today, however, it is increasingly the case that «own» data is used to estimate in advance how the official data will develop, and when the official figures do appear, positions have already been taken once or thrice. As examples, and as discussed previously in these pages, one may estimate consumer activities through real time drone data covering the number of cars parked outside shopping malls. Trade activities may be estimated by using real time and global satellite surveillance of which ships are and go where. Consumer satisfaction with Apple may be measured by AI and analysis of social media comments. And the list just goes on. The concept of “forecasting” is being replaced by the idea of “nowcasting”. The only remaining question is really whether an asset manager has access to, and the capability to manage, really big data.
Investors chase investment returns. Returns may come from beta, which is what the market gives you for free, and alpha that comes from beating the market. Beta is available from cheap or free (yeah, right…) index funds, and alpha can be bought from the best and the more expensive active managers. A special, but large group, are those who insist on buying what is really beta from expensive managers that were supposed to deliver alpha, but that is just silly. However, today even alpha is just not alpha. Using AI, learning models and big data processing, it is perfectly possible to identify systematic sources of alpha that can be easily replicated, such as long / short, carry, value and momentum strategies. These «false» alpha sources should not be too expensive to access, and there is an expanding universe of managers and funds providing systematic alpha at sensible prices. That leaves a pretty short list of managers that actually both charge for and provide real alpha. And the rest, who neither deliver beta at beta prices nor alpha at any price, should probably try fairly smartly to swap some of the soothsayers for members of the hoodies, Converse and portables-covered-with-weird-stickers legion. While they still have some assets left to manage.