Why the READ way is the better way
A non-academic article about the current practices applied to determine student loan interest rates and why they are wrong
Introduction
In this article, I explain why the READ model is not based on interest rates and credit scores. At the core, I describe why the commonly applied concept of interest rate is irrelevant to determine the real economic risk of capital for young borrowers.
To do so, I challenge two big financial ideas; first the currently prevalent theory of money creation and second the concept of credit scores. I argue that both are not ‘good’ indicators to determine the price of a student loan.
For the benefit of all readers, I attempt to build up my argumentation step by step.
Interest Rates in general
Interest rates simply put are nothing else, but the price of money for a specific use case and that price is derived from the risks associated with it, so that you could say, the price of money equals the risk of money.
Therefore, like any other ‘money purchase,’ in the case of student loans, the interest rate paid by a borrower is equal to the risk of that borrower to default, and this is commonly calculated from a risk-free interest rate and an individually determined risk premium based on credit scores.
To determine if the interest rate applied to a student loan ‘makes sense’ and reflects the borrower’s risk, we need to be able to answer the following two questions:
- What is the risk of default?
- What is the risk-free interest rate?
Default Risk
The risk of default occurs if a borrower is unable to meet his or her payment obligation as per the terms of the loan agreement.
The repayment includes two components; one related to the nominal loan amount and one associated with the interest rate applied to that loan amount.
In practice, we see that by default lenders calculate the payment obligation in the form of an annuity payment — a fixed amount for the duration of the contract period that includes both nominal repayment and interest rate.
The risk of default can be divided into two categories (I am assuming that there are no substantial savings available to tap into and that the loan is unsecured, i.e., there is no collateral or guarantee):
First, there is the problem of borrowing too much; if the loan amount is disproportionate to the disposable income, the borrower cannot afford the amount of debt given his salary and cost of living. This problem is fundamental. If the ratio is out of whack, and unless the future earnings potential is such that over time, the ratio is improving, there is frankly speaking little hope.
Second, there is the issue of a particular risk event occurring, this is the case if the income declines either temporarily or entirely, and therefore, not enough money is available to meet the required payment obligations. Examples for such risk events include underemployment, unemployment, disability or death.
Unlike fundamental default risks, secondary risks can be calculated accurately and therefore it is possible to mitigate them for example through insurance.
So why is this relevant? Logic (at least my logic) would argue that the common practice applied by 99% of lenders to determine the default risk on the basis of adding an individual risk premium based on credit scores to a risk-free interest rate only makes sense, if the risk-free interest rate and the credit score capture the described fundamental and secondary risk factors.
Our research has led us to the conclusion that this approach is flawed and shows weak links to the real risks we see in the student loan market.
In the following paragraphs, I will elaborate on each aspect in more detail.
Student loan interest rates & the economy
Let’s start with looking at how student loan interest rates have developed compared to some of the most commonly considered economic indicators: unemployment, inflation, and GDP growth.
Why do we look at unemployment? As outlined above unemployment is a secondary risk factor to determine default risk, the higher the unemployment rate, the higher the chance that a borrower cannot meet his or her payment obligations.
Why do we look at inflation? To determine the ratio between debt-to-income, relevant to the READ model is the development of the disposable income over time as this will change the ratio. As a proxy, we assume that the cost of living will grow at a specific rate of price inflation. One would hope that earnings increase accordingly at the very least matching the rate of price inflation. We know the economics behind this, unfortunately, don’t work quite like that, and I will elaborate on this next time, but for the purpose of this discussion, let’s park this topic for now.
Why do we look at GDP growth? It is essential to know where we are in the economic cycle and where we are heading to make meaningful predictions relevant to our model.
As an example, if the economy does well, companies should see higher demand for their services and products, therefore requiring more people to join the firm to meet the increasing demand, which in turn should result in higher profits, which can be distributed to employees.
Therefore, in theory during boom times, unemployment should be low, and income should increase. How income is distributed is a topic for another time, too, so more to that later.
We looked at data from 1992 to 2017, and this is what we have found:
Over the entire 15-year observation period average Fed Rates, Unemployment Rates, and CPI Growth Rates have been around the 2.5% mark each. Stafford interest rates on average were 3% higher than the Fed Rates.
Up until the Financial Crisis, Stafford rates (used as a proxy for the overall level of student loan interest rates) have nicely followed the development of Fed Rates.
After 2007 however, the gab has increased as we entered for the second time in the 20th-century into an era of Quantitative Easing, and as a consequence, the correlation between interest rates and other economic indicators has weakened substantially.
In fact, post-2007 the correlations between Fed Rates and GDP growth, unemployment and price inflation have all declined, and we would consider them as weak.
Because of that, we, therefore, must conclude that the level of interest rates is not a good measure to predict inflation, unemployment and growth rates — at least not for the last decade or so — and as such the interest rate level is not a ‘meaningful’ determining factor to the risks associated with student debt.
Risk-free interest rates
We then went one step further and asked the question of how the price of capital is determined. As the commonly used starting point is the risk-free interest rate, let’s start with that.
The risk-free interest rate is the rate of return for an investment with no risk of financial loss. It is common market practice that most financial institutions assume no government default and take the yield of a government-issued note or bond as a proxy.
Of course, a fundamental issue here is that of printing money to meet the payment obligation, resulting in a loss of purchasing power and therefore value.
Risk-free interest rates are widely applied in portfolio theory which is based on capital asset pricing models and as such a crucial component in determining the cost of capital and accordingly the risk of capital. This is also the case for student loan lenders, and for that matter, any consumer credit.
The economic schoolbook theory explanation goes along the lines of the following; market participants can borrow capital at the risk-free rate and therefore it is the starting point to determine the price of capital.
Arguably, however, in reality, we know this is seldom the case, but be that as it may for now. In short, the risk-free interest rate is considered as a reference to determine the base-line, non-diversifiable market risk.
So far so good. The next question then is, how is money created in the first place? Economic theory suggests that the price of a good or service is determined by supply and demand. However, how does that work exactly in the case of money?
Money creation
To answer this question, let’s start with looking at the three main theories currently recognized by academic experts.
The first and currently predominant theory is the financial intermediation theory, which states that banks take deposits and use the money received to provide credits.
The second is the fractional reserve theory that argues that banks do not create money, but the banking system as a whole is creating money.
The oldest theory is the one of credit creation that says that banks create money.
So, which one is it? The implications of which one is right might have profound consequences to policy-making and therefore the economy as a whole, it is, therefore, essential to understand how this works.
Professor Richard Werner has published an article back in 2015 that has addressed precisely this question. You might be surprised to know that leading economists have largely been ignoring banks and their role in the economy in their theories.
In an empirical test, Werner has simulated taking out an actual bank loan and has analyzed how the bank grants the credit from an accounting perspective.
Werner has proven that banks create “money out of nothing” and hence confirmed that the credit creation theory is correct. This result supports that both the currently prevailing financial intermediation theory, as well as the fractional reserve theory, are flawed, which leads to the obvious question of why market participants should use Fed-rates as the basis to determine risk-free interest rates when it is the banks who are responsible for the creation of money and therefore supply?
In our view, it is therefore questionable if the level of interest rate set by the Government is connected to real economic activities.
Given recent history, the financial system as it is currently set up has failed to assess risks appropriately, which in turn created several asset bubbles that led to the global financial crisis.
The findings of Werner, therefore, support our view that interest rates are not an accurate indicator to determine the real risks associated with student loans.
I highly recommend Werner’s paper for those of you who are interested to learn more about his empirical test. Here is a link to his article:
https://www.sciencedirect.com/science/article/pii/S1057521915001477?via%3Dihub
Credit scores
The last piece of the puzzle is credit scores. Lenders use credit scores to evaluate the potential risk posed by lending money to a borrower as to mitigate losses due to default risk, i.e., they determine who qualifies for a loan, at what interest rate, and what credit limits.
At first glance this seems to be a reasonable approach, but can it be applied to young graduates and professionals with student debt? To answer that question, we need to look at how credit scores are calculated.
In the United States, the credit score of a borrower is primarily based on credit report information, typically from one of the three major credit bureaus: Experian, TransUnion, and Equifax. Please note that neither current earnings nor employment history is considered when calculating credit scores.
FICO is the most widely used credit score system in the US. The generic or classic FICO credit score ranges between 300 and 850; the higher the score, the less likely he or she will go 90 days past due in the subsequent 24 months after the score has been calculated, i.e., the lower the risk of default and therefore the lower the risk premium.
According to the company’s website, 90 percent of all lending decisions in the U.S. use FICO scores, and more than 27 million scores are sold each day. So how is it calculated?
The payment history makes up for 35 percent of the total credit score and is the most important factor in calculating credit scores. According to FICO, past long-term behavior is used to forecast future long-term behavior. In the case of graduates and young professionals, it is impossible to have a ‘past long-term’ history, and therefore we consider 35% of the score as not applicable.
The second most crucial factor is credit utilization or the percentage of available credit that has been borrowed. It makes up 30 percent of the credit score. Again, we find this irrelevant for graduates and young professionals. By default, they will have little credit history and low limits, and so it comes as no surprise that their credit utilization is pretty high. Furthermore, this is also where the FICO system can be tricked. If one applies for several credit cards, the limit will increase, and credit utilization will come down, improving the overall credit score.
Next up is the length of credit history, accounting for 15 percent of the total credit score. It is evident to us that a fresh graduate cannot possibly do well here; he or she simply hasn’t lived long enough to build a relevant history.
Credit mix makes up another 10 percent of your score. While this is somewhat of a vague category, experts say that repaying a variety of debt products indicates the borrower can handle all sorts of credit. While we see how this might make sense for a 45-year-old, we find it difficult to apply this to a graduate or young professional.
For the sake of completeness, the remaining 10 percent of the score is made up of new credit.
Based on the above we find that 80% of the FICO score is irrelevant for young student loan borrowers; first, because it emphasizes historical data as opposed to the future potential, second because it disregards essential factors such as income level.
Conclusion
We think that the way traditional lenders currently determine the interest rate for an individual student loan borrower is in principal flawed.
First, because the assessment of the base risk, i.e., the risk-free interest rate, is not connected to real economic risk and second, the way risk premiums are determined are not forward-looking, which makes not much sense for our target audience.
At READ we, therefore, disregard both the level of risk-free interest rates as well as credit scores. Instead, we look at each individual case separately and tailor our refinancing terms accordingly.
Disclosure
The opinions expressed in this article are mine. My conclusions are derived from both my own primary research as well as 3rd party secondary data.
About READ
READ Enterprises has developed an alternative risk assessment approach to the traditional lending industry.
Visit our website https://readenterprises.org/, and follow me on twitter https://twitter.com/AlbiRead
