Heuristic Bias and the Forex Markets

The first question asked to me as a forecaster by a foreign investor few years back was not on Gold or Oil or India, but on Euro Dollar Forex rate. Such is the engagement with Forex that internationally Forex is the top traded asset. And after the equity preoccupation has reduced with the equity slowdown, this engagement with Forex has only increased. We see online trading platforms for Forex spurting out each day, enticing investors to put a few 100 dollars and allowing them to trade with a leverage of 100. A 500 dollar put in by a small investor can play up to 50,000 dollars of exposure. The preoccupation or excessive speculation in the Forex market is not just owing to these daring day traders. But also because of the normal investor on the street, who is using a different kind of leverage, the credit leverage. The idea that lending Euros is easier as it comes with a 3-5% interest rate compared to say the Romanian Lei, with a double digit rate, invariably pushes the investor to borrow in Euro.

This heuristic bias, or thumb rules are also extended to the fact that Euro is inherently considered stronger than many emerging market currencies, so a loan in Euro is perceived better than that on a local emerging currency. Be it the real estate owners, tenants, builders, or simple personal finance purchasers, the Euro bias is a part of the large global credit character. Moreover, even if one is not a trader, there is a monthly car installment or mortgage to pay, less or more is the preoccupation that increases and decreases as the rate fluctuates. A local currency exposure is less adventurous and exciting when you have something international or global to speculate on.
Similar cases have happened in other economies around the world like Indonesia and Thailand before the South East Asian crisis when two-thirds of the region companies had 40% of their debt in foreign currencies. The text book case of interest rate arbitrage traps a majority of the Forex users who fail to see beyond the interest rate differential. What happened after the South East Asian crisis was an 80% collapses in currency values in Asia and the collapse of the Turkish Lira in 2001, old stories from econohistory. The excessive speculation in Forex happens both in normal and extraordinary times.
A search on the Turkish crisis in 2001 will take you all over the information loop regarding the political crisis, the deficit, inflation etc. And it’s a rare research blog or institution that will illustrate the steps that led to the currency crisis. What happened in Turkey too was also the classic case of Turkish Banks performing interest rate arbitrage. This positioned the local bank’s profitability in the foreign hard currency. The arbitrage activity could have continued as long as the exchange rate at the end of the period did not change radically against the banks. A severe devaluation, which was not expected happened and forced the lira 36% down in two days.
These cases have a limitation when we look at the Indian subcontinent owing to “convertibility issues” or to the very fact that we still have not got used to trading currency as masses. This has its advantages. But that’s changing, as India learns to speculate on the Rupee Futures. We also have a large corporate trading based on forwards. Forward bias is another illusion. Ideally and conventional wisdom might suggest that a forward premium in the Forex market would be indicative of direction the currency might take in the coming trading days. But research has proved that forward rates are a poor predictor of future rates. No wonder surveys have shown traders betting against the forward rate. Even futures premiums and discounts on equity side perform dismally as predictive tools. Hersh Shefrin in his book ‘ Beyond Greed and Fear’ raises the question about profiting from betting against the forward rate. He proves that the prediction of error tends to move in cycles. When the error is positive, it tends to stay positive for a while before turning negative. Once negative, it stays negative for a while. Forex traders overreact, bet on trends and are overconfident making Forex market as inefficient as any other asset class.
According to Shefrin, it’s not greed and fear, but hope and fear. We all commit forecasting or investing errors because a majority of us rely on rules of thumb (heuristics). Conventional research is primarily extrapolation. We don’t live in a rational environment i.e. the world we live in is not error free. This is why the most exciting of riskless arbitrages have brought economies and companies like LTCM (Long Term Capital Management) crashing. This is why conventional Long and Short funds have failed, despite being market neutral. In a research paper published in the July 1985 issue of the Journal of Finance, De Bondt and Richard Thaler, argue that investors overreact to both bad news and good news. Therefore, overreaction leads to past losers becoming underpriced and past winners becoming overpriced. Heuristic biases are ubiquitous, germane and very expensive.
Shefrin classifies the Heuristic biases in his book. Availability bias is like the first recall, what comes in mind first seems right. Representativeness, suggests investors become unduly pessimistic about prospects of past losers, which turn out to give exceptional returns. Regression to mean is a bias too, as historical average return on equity does not mean much for equity performance in the near future. Historical averages or regression to mean is a bias with little predictive ability. Overconfidence and accuracy generally does not go hand in hand. When people are overconfident, they set overly narrow confidence bands and end up being “humbled” and surprised. Aversion to ambiguity bias explains why people prefer the familiar to the unfamiliar. Bailouts don’t happen to save the world from the crisis, but more because people are scared of fathoming the depth of the unfamiliar ground. Emotion and cognition…first we did not expect the markets in India to crash, now that they have crashed a majority of the investors might spend years worrying about it that it may happen again.
Heuristic biases are also the reasons why conventional research gets surprised. Past performance is the best indicator of future performance is a heuristic error, generally imperfect. The best way to analyze extrapolation and accountability of conventional research is to look at inflexion points. Look at Yen Dollar rate and press releases issued by the Bank of Japan and the top researchers in JAN 2007 when the pair was ruling at 122 and this is what news the search might get…“In Tokyo, the yen dropped following the BOJ’s rate decision to leave interest rates unchanged” This as we know was followed by the historic strengthening of the yen against the dollar. And now change YEN with INR and look at 15 Nov 2007 Rupee levels at 39 and hit search and you might get a Reuters Poll on the same date. Axis, Bank of America, Bank of Baroda, Bank of Nova Scotia, BNP Paribas, Calyon, Citigroup, DBS, Deutsche Bank, HDFC Bank, ING Vysya Bank, JP Morgan, Kotak Mahindra, Lehman, NCDEX, Royal bank of Scotland, Standard Chartered Bank, State Bank of India and Syndicate Bank participated in poll. All of them were forecasting INR for DEC 2008. They were nearly in consensus of where the rupee should go after a year. The Median was 37.63 (highest 41 – lowest 36.6). What happened was 50.58. We at Orpheus are not free from heuristic errors and bias. But we are trying.