국민연금 거래의 단기 정보력에 대한 연구 (A Study on the Short-Term Informativeness of Transactions by the National Pension Fund)

서지정보

· 발행기관 : 한국재무학회
· 수록지 정보 : 재무연구 / 32권 / 4호 / 651 ~ 689페이지
· 저자명 : 양철원, 우민철

초록

본 연구는 국민연금 거래와 이후 주가수익률과의 관계를 분석함으로써 국민연금 거래가 지닌 단기 정보력을 검증하였다. 실증분석을 위해 국민연금의 일별 순매수금액 자료를 사용하였으며, 방법론으로 재무에서 전통적으로 사용하는 포트폴리오 접근법과 회귀분석을 사용하였다. 1일 보유수익률 분석 결과, 포트폴리오 방법과 회귀분석 모두에서 국민연금 순매수금액비율이 높은 포트폴리오가 더 높은 수익률을 보여주었다. 순매수금액비율이 가장 낮은 포트폴리오를 공매도하고 가장 높은 포트폴리오를 매입하는 헤지포트폴리오의 수익률이 일평균 1.24%로 유의미하였다. CAPM과 Fama and French(1993)의 3요인 모형을 사용하여 위험을 조정한 후의 수익률을 살펴보았을 때도 동일한 결과를 확인하였다. 하지만 보유기간을 늘렸을 때, 3일 이후의 보유수익률부터는 헤지포트폴리오의 수익률이 유의미하지 않았으며, 4~5일에는 음(-)의 수익률로 반전되었다. 위의 결과들을 설명할 수 있는 여러 가설들을 검토하였을 때, 본 연구결과가 국민연금이 단기적 사적정보(private information)를 가지고 있음을 의미하기 보다는 국민연금의 대규모 거래로 인한 일시적인 가격압박(price pressure)이 존재함을 지지하는 결과로 해석하였다.

영어초록

This study verifies the short-term informativeness of the National Pension Fund transaction. It also tests whether the National Pension Fund is an informed trader with meaningful private information. There may be contradictory views on this. The National Pension Fund has specialized management personnel with analytical and investment skills that go beyond ordinary people in stock investment in connection with stock management. In addition, the National Pension Fund has a special status with companies in the Korean stock market, which makes it easier for the fund to access insider information than it is for other investors. Because of this, transactions in the National Pension Fund can have information power. However, the National Pension Fund is constrained by various regulations governing active investment in risky assets and is exposed to political influences that may hinder investment efficiency. These factors may make the trading of the National Pension Fund less informative. This problem is examined through empirical analysis.
We calculate the daily net purchase of the National Pension Fund using the transaction data provided by the Korea Exchange (KRX). Empirical analysis of stock price predictability is based on the portfolio approach and regression analysis, which are traditionally used in finance. The empirical analysis uses the daily net purchase amount of the National Pension Fund scaled by daily trading volume 1 to 5 days after the trading day. The results of monthly observations are also analyzed through the robustness analysis.
The main results of the study are as follows. First, the stock price predictability of National Pension Fund transactions is verified using the portfolio sorting method. When one day’s holdings of returns are calculated, portfolios with a higher net purchase of the National Pension Fund have higher returns than those with a lower portfolio [Check whether this should be “lower net purchase.”]. The return on hedged portfolios, which buy the portfolio with the highest net buying value and sell the lowest portfolio, is 1.24% per day with statistical significance. The same results are obtained when we examine the returns after adjusting the risk using the CAPM (capital asset pricing model) and the three-factor model of Fama and French (1993). When the holding period is increased, the statistical significance disappears after 3 days, and the return becomes negative after 4 days.
Second, we examine the short-term informativeness of the National Pension Fund transaction according to the characteristics of firms. Empirical results show that stock price predictability is more concentrated on small stocks.
A robustness test using regression analysis shows the same results. The regression coefficient is significant when regression analysis is performed using the stock returns as a dependent variable and the net purchase of National Pension Fund as an independent variable. The significance is maintained when other control variables are added.
The above results show that National Pension Fund transactions are informative but short-term. The results can be interpreted in two ways. First, the stock price prediction power may be due to short-term (about 1–2 days) private information of the National Pension Fund. There are two ways the National Pension Fund can acquire information. First, it is possible that the National Pension Fund will acquire the inside information of the company using its dominant position as a major shareholder and then deal with it one to two days before the public announcement. However, it is difficult to accept that the National Pension Fund is continuously carrying out such risky illegal insider trading. Another possibility is that the National Pension Fund has an excellent ability to interpret short-term stock price trends. Previous studies report that the National Pension Fund takes a contrarian strategy for the market and individual stock conditions. When the stock price is excessively lower than its fundamental value, the National Pension Fund can buy stocks, and the price may rise afterward. In the same way, when the stock price has risen excessively, the National Pension Fund can sell stocks and make a profit. However, the National Pension Fund has introduced a long-term method of allocating assets rather than investing in a daily portfolio adjustment scheme. In light of these points, the explanation based on short-term private information of the National Pension Fund is not appropriate.
The second interpretation is that the predictive power of stock prices is due to the temporary increase in price due to National Pension Fund transactions. There may be price pressures if other traders follow the transactions of the National Pension Fund. Woo and Kim (2018) analyze the possibility that the National Pension Fund transaction could indirectly affect the market by influencing other investors. They find that institutional investors, especially investment trusts and other financial institutions, trade in line with the net buying of the National Pension Fund. Based on these results, it seems reasonable to interpret the empirical findings in this paper as a result of temporary price pressure.

참고자료

· 없음