The advancement in digital technology provides room for innovations in many sectors, including in the financial sector. The development of digital currencies is one example of digital advancements in this sector. Digital currencies can take many forms, and cryptocurrency is one of the most popular forms of digital currencies today. Bitcoin and other cryptocurrencies (or altcoins) have slowly been adopted around the world as a means of payment. However, Bitcoin and altcoins are not issued by the central banks or other central authorities, making it difficult to regulate. At the same time, in several countries, the use of cash is declining to a very low rate. As the response to these economic phenomena, central banks in some countries are investigating the possibility of introducing their own version of digital currency which is called central bank-issued digital currency or CBDC. This research shows how to implement transaction cost economics (TCE) framework to characterize transaction costs in several CBDC design options through in-depth semi-structured interview sessions. The interviewees were the experts from central banks, commercial bank, regulatory bodies, and relevant NGOs in the European developed economies. The results of the interview sessions showed several findings. First, according to the experts, there are three main CBDC design options: Retail CBDC, Wholesale CBDC, and Full-reserve Depository Banks CBDC. Second, among the three transactional dimensions of TCE (uncertainty, transaction frequency, and asset specificity), the uncertainty aspect provides a significant contribution to overall transaction costs. Third, the central bank and commercial banks are the ones that bear most of the transaction costs, especially in Retail CBDC design option, while the households and non-financial institutions do not need to bear any transaction costs.

In this thesis, the System Dynamics (SD) model of the Full Reserve Banking (FRB) system built by Van Egmond and De Vries (2016) has been reviewed and improved. The improved model is used to see the effect of FRB system in preventing a long-lasting recession as one occurred in 2008. FRB system is a monetary reform which attempts to separate private banks’ monetary and credit functions. Under FRB system, central bank creates money depending on the price level. Therefore, theoretically, FRB system can prevent recessions by maintaining the price level constant. Van Egmond and De Vries have modeled FRB system using System Dynamics (SD) modeling technique. However, their model has a few model problems which reduce reliability of the model; Ⅰ) the model crashes in different conditions, Ⅱ) the model is based on some assumptions that are less realistic or logical Ⅲ) the measurement units of the model are not specified. In this thesis the solutions for these model problems were sought through a model review and desk research. In the original model, high house prices and debt level are a direct and definite reason why the recession begins. Households default because house prices went up too fast and they could not afford the debt repayment. However, in the improved model, households default because their wages are stagnated. Households default when wages are stagnated while debt level grows high. In both original model and improved model, FRB system successfully protects the price and wage level. Under FRB system, the probability of recession decreases significantly. However FRB system is found to be not effective in stabilizing the price level when the firms’ dividend payout ratio is high. Furthermore, FRB system was not necessary as households do not default when the firms’ dividend payout ratio is low. The reason that FRB system’s effectiveness is decided by dividend payout ratio is because the model was calibrated originally with dividend payout ratio of 100% which is unrealistic in the real world. Therefore, the model is expected to be calibrated properly especially focusing on dividend payout ratio in a future work. Furthermore, it is recommended to include trades with other countries in the model in further work.