There are many definitions of arbitration. Arbitration is generally perceived as a trading operation that uses the prices of assets in different markets at a certain point in time to generate risk-free profits. Professor Aswat Damodaran of the Stern School of Business at the University of New York distinguishes between three types of arbitration transactions: pure arbitration, "exemplary" arbitration, and speculative arbitration. In particular, net arbitration involves investments without investing their funds, without any risk and with net profit.
In pure arbitration, a market participant does not risk anything but makes a profit above the risk-free interest rate. A "rough" arbitration implies that there are assets that have the same or almost identical cash flows, but different prices. Speculative arbitrage, which cannot be called a full-fledged arbitration, involves market participants looking for undervalued or overvalued assets and similar assets by buying cheap assets and selling expensive ones.
It is generally accepted that financial markets do not provide risk-free arbitration opportunities, even taking into account transaction costs. While it may be assumed in some financial contexts that there are no arbitrage possibilities, there are still rational reasons for explaining the existence of financial market arbitrations. Such opportunities will be discussed in this paper. In general, the postulate of a lack of arbitration opportunities dates back to the so-called "paradox", which C wrote about for the first time. Grossman and J. Stiglitz. Thus, in 1976 and 1980 Grossman and Stiglitz published the theory that in an efficient market all information is already taken into account in the current prices of financial assets, and, consequently, no one is motivated to obtain information that affects prices. In addition, if arbitration opportunities arise, they are quickly exhausted by economic agents.
Besides, the theory of market microstructure shows how price differences can occur for identical assets in markets that are not fully centralized, segmented, or not entirely transparent. At the same time, repeated empirical studies have failed to detect short-term (intraday) arbitrage opportunities in a variety of markets. Given the high level of activity in major financial markets, such short-term arbitrage opportunities can only be adequately explored using real-time quotes of all instruments involved in the arbitration.
All relevant aspects of the market microstructure should be taken into account in order to take into account not only profit-making opportunities but also the transaction costs faced by participants in the market. For this reason, this paper will not address intraday arbitration. It often turns out that it is impossible to make two short-term intraday transactions simultaneously, or technically impossible, or difficult to get the arbitration profit in the conditions of a high rate of change of quotations. In particular, this problem was studied even in the Federal Reserve Bank of New York.
Thus, Ernst Schaumburg, a researcher at the Federal Reserve Bank of New York, found out that the growth of high-frequency trading robots and the development of algorithmic trading, in general, may contribute to the increase of efficiency in the currency market. Under such conditions, it becomes more and more difficult to make arbitrage profit on intraday quotations. The share of high-frequency trading (HFT) has increased by 25% today, having risen from almost zero compared to the beginning of the 2000s, according to the Bank for International Settlements. In his paper, Schaumburg gave an example in which an investor could use 1 billion euros to buy $1.316 billion, then convert those funds to 154.2 billion yen, and then finally exchange it again for the single European currency of Europe, earning a return of 120.65 euros for each invested $1 million. The so-called round trip, or triangular arbitration, does not include transaction costs.
Transaction costs are a key factor in determining the economic viability of an arbitration transaction. In particular, the point is that any arbitration algorithm must take into account transaction costs, execution risks, and other details that are often overlooked in theoretical works. When there are undervalued or overvalued assets in the market, this does not necessarily mean that risk-free profits can be guaranteed. This is due to the fact that prices for financial instruments often move within the framework of the so-called "no arbitrage zones". These zones are price intervals at which it is impossible to derive arbitrage profits due to transaction costs and any other barriers. The higher the transaction costs, the larger the arbitration zone. Thus, arbitration strategies can only be implemented when prices move out of the arbitration zones