“My dad is stronger than your dad.” “Our car is faster than your car.”

Remember when you were a kid and had to listen to all those “my stuff is cooler than your stuff” arguments?

Well, those arguments are still happening in the wealth management industry today – only this time they’re around the pros and cons of firms’ algorithms.

Algorithms have become a feature of the wealth management landscape, playing a vital role in how clients’ money is managed. But which algo will best satisfy clients’ investment priorities and service needs?

In many ways, we are still caught in the “my dad is stronger than your dad” game. We are surrounded by algorithms. Yet as end consumers of products and services, we have limited facts and insight into which will provide the best outcome.

And as a wealth or asset manager that has invested heavily in building algorithms, how can you determine if yours are better than your competitors’? Can we even measure the success of one algo versus another, and do we really need to?

Success depends on data

What is certain is that algos need data. Accurate data, and lots of it.

Algorithms will never be better than the amount and quality of the data that feeds them. It’s as simple and complex as that. You can put all the science, brains and money available into creating sophisticated, ground-breaking algos, but without relevant, high-quality data they are useless.

Data has never been more available, so getting it should be easy, right? The answer is yes … to a degree. The more important question is what data you get, how you get it and what you do with it.

I like to think there are five ground rules when it comes to algos and data:

1. It has to be accurate. As the saying goes: “garbage in, garbage out.” Data quality and governance is key to having an algo that will deliver meaningful support for your business.
2. It needs to be relevant. You can’t just fill up your database with whatever you find. That data will contradict and conflict with each other, polluting the input first and output second. The right data is what counts.
3. The data should be unique where possible. If you want your algorithms to help differentiate you from your competitors, this could be the most important factor.
4. The algo needs to capture and process data in an efficient manner. Data should be available through APIs and stored in a way that makes it easy for the algo to leverage.
5. Poor data also means a poor user experience, and loss of trust in a business where you simply cannot survive without it.

If you don’t have control over your data, and are not able to manage it in an efficient way, algo quality will be affected. Since your data often resides in your core systems, these systems will be vital for your algorithms—and hence the efficiency and success of your business.

If you don’t have an algo in your direct environment today, it’s likely you will soon. So doing the necessary prep work is vital.

Do you have a healthy environment for your algo to live in? Do you know the current quality of your data? What are you doing to improve it? Is the data captured and stored in a way that makes it easily accessible? Is your platform ready for the ever-increasing need for data and volumes?

Algos, AI and machine learning may make your firm look cool, but their practical effectiveness starts with the basics. That means getting your core systems running smoothly with high-quality data. For at the end of the day, cool stuff that breaks is no longer cool. And you don’t want to be the loser in that argument again, do you?