Imagine a giant, living city where every house is a character in a massive, chaotic story. This isn't just about bricks and mortar; it's about the invisible financial ropes tying people to their homes. In the world of economics, scientists often try to understand these stories by looking at the "average" person, as if everyone were identical twins living in identical houses. But real life is messy. People have different jobs, different savings, and live in neighborhoods that rise and fall in value at different speeds. To understand how families might lose their homes or pay off their loans early, researchers use a special tool called an "Agent-Based Model." Think of this as a super-advanced video game simulation where thousands of unique digital characters (agents) make their own decisions based on their own money, their own neighborhood, and the rules of the bank. By watching this digital city play out, scientists can see how small changes in interest rates or house prices can cause a ripple effect, turning a few struggling families into a city-wide crisis. This matters because when too many people can't pay their mortgages, it doesn't just hurt them; it can shake the entire economy, making it harder for anyone to get a loan or buy a home.

The paper you're about to read dives into this digital city, but with a very specific setting: Ankara, Turkey. The authors, Bilgi Yilmaz and Berna N. Yilmaz, built a custom simulation to figure out why some homeowners in this city are more likely to lose their houses than others. They didn't just look at the people; they looked at the places. They realized that a house in a fancy, central district behaves very differently from a house in a far-out suburb, even if the mortgage paperwork looks the same. They also had to account for the unique rules of the Turkish banking system, where the government sets strict limits on how much money you can borrow compared to your house's value, and where the law allows banks to chase you for your other assets if you stop paying.

Here is the story of what they found, told through the lens of their digital experiment.

The Setup: A City of Digital Households

The researchers created a virtual Ankara with 3,200 digital households. Each household was given a unique profile: a specific income that grows (or shrinks) over time, a specific house in one of ten different districts, and a mortgage. The simulation ran for five years (60 months) in the digital world.

To make it realistic, they fed the model real data from 11,092 actual house listings in Ankara. This allowed them to map out exactly how much houses cost in different neighborhoods. They found that a house in the central district of Çankaya was worth nearly twice as much as a similar-sized house in the district of Sincan. This "spatial sorting" meant that where you lived was just as important as how much money you made.

They also programmed the model with the reality of Turkish interest rates. In recent years, these rates have been incredibly high, hovering around 39% per year. In the simulation, this created a unique problem: because the interest rate was so high, the monthly payments went almost entirely to paying off the interest, not the actual loan amount. It was like trying to fill a bathtub with a leaky hose; you're pouring water in, but the hole (interest) is so big the water level (your equity) barely rises.

The Rules of the Game

The digital homeowners had two main choices to make during the simulation:

1. Refinance: If interest rates dropped, they could try to swap their expensive loan for a cheaper one. But there was a catch: the bank would only say "yes" if the homeowner had built up enough equity (at least 15% of the house's value). If the house value dropped too low, the bank would slam the door shut, no matter how much money the homeowner wanted to save.
2. Default: If the house lost value and the homeowner couldn't pay, they could walk away. But in Turkey, walking away isn't free. Because of "full recourse" laws, the bank could come after the homeowner's other savings or assets. This made the decision to default a scary, high-stakes choice, not just a simple math problem.

The Great Crash Simulation

To test the system, the researchers introduced a sudden, nasty shock at the 24-month mark (two years in). They simulated a 14% drop in house prices across the city. In the real world, this is like a sudden economic storm that makes everyone's home worth less overnight.

Here is where the story gets interesting. The authors found that the "average" homeowner was actually quite safe. Even with that 14% price drop, the median borrower still had enough equity buffer to keep paying. The simulation showed that the slow amortization (the fact that they weren't paying down the principal fast) combined with the high initial house values meant that a moderate price drop wasn't enough to break them.

However, the story changed completely for the high-risk group.

The researchers discovered that the danger wasn't spread evenly across the city. It was concentrated in the hands of people who started with the highest loans. They looked at the borrowers in the top quartile (the top 25%) of loan-to-value ratios—those who borrowed the most money relative to their home's value. When the price shock hit, this group didn't just struggle; they collapsed.

In the simulation, when the house prices dropped, the high-leverage borrowers saw their equity vanish instantly. Because they had borrowed so much to begin with, a small drop in price pushed them deep into "negative equity" (owing more than the house is worth). Since they also had high monthly payments relative to their income, they hit a "dual trigger": they were underwater and they couldn't afford the payments.

The Three Big Discoveries

The simulation revealed three main lessons about how this digital city works:

1. High Interest Rates Create a "Slow-Start" Trap
Because the interest rates were so high (around 39% APR), the borrowers were paying off their loans incredibly slowly. In the first five years, the average borrower had only paid off about 13% of their actual loan balance. This meant that for a long time, their "safety net" (equity) was very thin. If house prices dropped, they had almost no cushion to absorb the blow. The paper suggests that in high-interest environments, the risk of default is less about the interest rate changing and more about the house price dropping, because the loan balance stays stubbornly high.

2. Location is Everything
The paper explicitly rules out the idea that all houses in a city react the same way. The simulation showed that because different districts grew at different speeds, two neighbors with identical mortgages could have totally different fates. A borrower in a fast-growing, high-value district might build equity quickly and be able to refinance. A borrower in a slower-growing, lower-value district might get stuck. The "spatial sorting" meant that financial vulnerability wasn't just about the person; it was about the neighborhood.

3. The "Refinance Wall"
This was a crucial finding. The simulation showed that many homeowners wanted to refinance to save money when rates dropped, but they couldn't. The bank's rule (a limit of 85% loan-to-value) acted like a wall. If a homeowner's house value dropped even a little, they hit this wall and were locked out of refinancing, even if it would have saved them thousands of dollars. The authors found that this regulatory barrier trapped many households in expensive loans, making them more likely to default when a price shock hit.

The Verdict: Who Gets Hurt?

When the dust settled after the simulated crash, the results were stark.

• The Safe Zone: Borrowers who started with lower loans (the bottom 75% of the group) were largely unharmed. They could refinance or keep paying.
• The Danger Zone: The top 25% of borrowers (those with the highest initial loans) were the ones who lost their homes. In the stress test, more than half of this group defaulted.

The paper suggests that the idea of "strategic default" (where people just walk away because they don't want to pay) is less common in this market than in places like the US. Why? Because of the "full recourse" rule. The digital homeowners knew that if they walked away, the bank could take their other assets. So, they only defaulted when they were truly broke and underwater, not just because they wanted to.

What This Means for the Real World

The authors conclude that we can't just look at the average person to understand mortgage risk. We have to look at the specific mix of high interest rates, strict bank rules, and local house prices. They found that in a high-interest environment, the system is fragile for those who borrowed the most. A small drop in house prices can push the most leveraged families over the edge, while the rest of the city remains standing.

The simulation also hinted that if the government or banks relaxed the strict "85% loan-to-value" rule, more people might be able to refinance and avoid default. But, as the paper notes, this is a delicate balance; loosening the rules too much might encourage people to borrow too much in the first place.

In the end, this digital story of Ankara teaches us that in a complex, high-interest world, your mortgage isn't just a contract with a bank. It's a tightrope walk where your balance depends on the neighborhood you live in, the speed at which you pay down your debt, and the invisible walls built by banking regulations. And for those walking the tightrope with the least amount of safety net, a single gust of wind can be enough to make them fall.