Can artificial intelligence use the 30-year mortgage rate to predict home prices?  It’s a question that has long fascinated the minds of every human.  Can we build a machine so intelligent that it could predict human markets?

One such machine is a decision tree.  Though it’s actually a model, it takes in multiple factors and produces a decision tree that predicts whether something is what you think it is.

For example, suppose you’re interested in predicting whether an image is a dog or a cat.  Artificial intelligence will give you a model’s guess.

So, can artificial intelligence use the 30-year mortgage rate to predict home prices?

Now, to the question at hand: predicting home prices based upon the 30-year mortgage rate.

An early-stage attempt follows.  Let’s start at the top.

At the top is Node 1.  Below the node is a choice.  If the 30-year mortgage rate is above 4%, then one answers “yes” to the question and follows the line to the next node (Node 2).  Since rates are now above 4%, we would follow this path.

Before following the current path, let’s take a moment and assume mortgage rates are below 4%.  In this case, the answer to the question is “no.”  This means following the tree to the right, which leads to terminal Node 3 (circled at the bottom right).  When rates are below 4%, one would expect home prices to grow by 6.5%, on average.  This is what we saw prior to 2018; rates were quite low, and how home prices were growing at a healthy pace.

Now, let’s switch back to where the world shifted to in 2018; mortgage rates well above 4% and heading potentially above 5% in 2019.

The question at Node 2 is whether the 30-year mortgage rate is above or below 5.3%.  Because the rate is below 5.3%, the answer is “yes”.  We therefore follow the tree to Node 4.

Node 4 asks whether the mortgage rate is above or below 4.4%.  For part of 2018, the average 30-year mortgage rate was below 4.4%.  We would therefore follow the tree to the right, which leads to a final Node 9.  Node 9 predicts an average home price appreciation of 3.5%.

Now, let’s shift to the latter half of 2018 when the 30-year was above 4.4%.  Following the tree to the left leads to Node 8.  This predicts home prices dropping by 4%.

What does this really mean?  Essentially, it means that housing prices are probably going to slow down in 2019.  They may not turn negative (in most places), but housing prices certainly won’t grow as much as they did in recent memory.

On the upside, passive income-producing homes might prove to be a better bet than the recent resurgence of a highly volatile stock market!

Conclusion: Perhaps Artificial Intelligence Can Use the 30-Year Mortgage Rate to Predict Home Prices

In looking at an early-stage artificial intelligence model that connects housing prices to changes in the 30-year mortgage rate, the 2019 year looks shaky right now.  The recent decline in mortgage rates has lowered the concern for the housing market.  However, there is still reason to be concerned.




Leave a Reply

Your email address will not be published. Required fields are marked *