Circular References in Financial Modelling
A lot of financial modeling takes place in Microsoft Excel. One of the errors that financial modelers come across during the financial modeling process is called the “Circular Reference” error. This error can affect many values in any model. To an untrained financial modeler, this could be the source of a lot of panic. However, […]
The Barnewall Model
The traditional financial theory assumes that all investors are rational. Hence, they believe that all investors will reach the exact same conclusion with regard to investment decisions. However, we see evidence of the opposite happening in the marketplace. Even if different people have the same information, they tend to process the information differently and come […]
Stock Valuation – The Discounted Cash Flow Approach in Detail
In the past article we have seen how Discounted Cash Flow (DCF) is the most appropriate method of stock valuation because it is rational and objective. Now, it is time we have a look at the details of this model. Present Value of Expected Future Cash Flows The basic of this model seems to be […]
Dividend Discount Model: Advantages
Dividend discount models are the first type of discounted cash flow models that we will study. The model simply discounts cash flows at a given rate just like any other DCF model. The difference lies in the fact that dividend discount models consider only “dividends” as being legitimate cash flows. Therefore, if a firm pays […]
Private Equity Investments in the Sports World
Sporting franchises across the world have traditionally been owned by wealthy individuals. This has largely been because of the fact that sporting franchises have been viewed less as financially viable investments and more as vanity toys. However, all this has started changing in the wake of stellar returns provided by sporting franchises in the recent […]
We have already seen the various ways in which technology has shaped the commercial banking domain. However, the commercial banking domain is so vast that it is being concurrently affected by many big changes.
The increasing use of artificial intelligence for commercial lending is another important trend that is impacting the commercial bank lending industry.
The availability of large databases where transactional data related to previous loans can be easily stored as well as the fall in the prices of computing power has made it possible to use big data techniques in commercial lending.
Artificial intelligence is often confused as being extremely complex software like the ones seen in sci-fi movies. In reality, artificial intelligence software is software that can harness data and analyze statistics at an astounding pace.
In this article, we will have a closer look at how artificial intelligence impacts commercial bank lending.
We will also try to understand the various benefits of commercial bank lending.
The ability to scale up a commercial banking loan portfolio without compromising on the quality of loans being issued has been a huge pain area for many commercial banks.
Commercial banks do want to give out loans to more corporations. However, they face a multitude of problems.
For instance, the lending process at commercial banks is dependent upon the judgment of individuals. Even though banks have tried to make these operations more process-driven, individual choices and opinions still play a huge role in the process.
Also, if a bank wants to expand its loan portfolio, it will have to hire more underwriters which ends up increasing the administrative costs.
Lastly, with the increase in the quantum of commercial lending, a dip in the quality of loans being made is commonly observed.
In order to overcome these problems, technology solutions driven by artificial intelligence and machine learning are being implemented on a large scale in the field of corporate lending.
There are two main types of artificial intelligence-based lending mechanisms which are being deployed in the field of commercial banking.
This means that the loan underwriting process is not completely machine-driven. Human intervention is required to explain to the computer the difference between a good and a bad loan.
Human intervention may also be required to help the software identify the various characteristics which differentiate a good loan from a bad one. Once the basic setup is done, the software will continue refining the model by incorporating more and more data that flows into the system.
The problem with supervised artificial intelligence-based lending is that it is still not completely rid of human bias because at the end of the day humans are feeding variables into a computer model.
The software uses a data mining technique to group loans based on common characteristics. It then evaluates the various patterns at play to differentiate a good-quality loan from a bad-quality one.
Once a new file is presented to the software for underwriting, it compares the characteristics of the new loan with that of the previous loans in order to help understand the risk-return profile of the loan.
The problem with unsupervised artificial intelligence is that this technology is still at a pretty nascent stage. Hence, many commercial banks are not comfortable taking large risks based on the decisions made by an algorithm. The technology will need to be around for a few more years before bankers show enough confidence in it.
The best part about artificial intelligence-based lending systems is that they are not very expensive to implement. Of course, bigger multinational banks will spend millions of dollars in capital expenditure in order to build proprietary technology which can then be leveraged for a long period of time.
However, it is not necessary for every bank to spend millions of dollars. In fact, the need for any kind of capital expenditure can be completely eliminated. There are many companies that are providing cloud-based solutions for artificial intelligence-based commercial lending.
Hence, it is quite plausible for any bank to adopt a pay-as-you-go system and focus on an operational expenditure-based system.
The flip side of having an operational expenditure-based system is that the bank may lose control over its proprietary data. However, given the number of software vendors which are providing plug-and-play systems, it is likely that the privacy concerns will also be soon taken care of.
The bottom line is that the rise of artificial intelligence in commercial lending is now imminent. Right now, commercial banks are using supervised artificial intelligence techniques. Over the years, it is likely that these same banks may transition towards unsupervised techniques.
Your email address will not be published. Required fields are marked *
