Appendix A: Indices
In this Appendix, we provide information about the three main stock indices on the Oslo Stock Exchange.
A.1 OBX
The OBX Total return index consists of 25 constituents. These constituents are the most liquid stocks available on the Oslo Stock Exchange. The liquidity measure is based on the last six months’ trading volume. The OBX index is adjusted for dividends and it is revised every six months. Several capping rules apply to the index. The largest component is not allowed to exceed 30 % of the total value. Remaining stocks are capped at a maximum 15 %, while non-EEA-stocks are set to a maximum 10 %. Between revising dates, the number of stocks of each index member are held constant. The OBX is a publicly traded index with both futures and options written with the OBX as an underlying instrument.
The index always has 25 index members. However, because of mergers, splits, reversed splits, revising dates, etc. we lack return observations for all 25 stocks for a few days in our sample. For the vast majority of dates, we have returns for all 25 stocks that are included in the index.
A.2 OSEBX
The Oslo Stock Exchange benchmark index is an investible index that consists of the most traded stocks on the Oslo Stock Exchange. It is revised twice per year and it is adjusted for dividend payments and other corporate actions. Between revising dates, the number of stocks of each security is fixed. Although not a rule, all OBX stocks are also part of the OSEBX. In other words, all the 25 most liquid stocks are always among the constituents of OSEBX.
The number of underlying instruments varies. Our estimation results use data from February 2009 through April 2013. In this period, the index has had between 53 and 61 underlying instruments. Our observed returns in the same period have been between 38 and 61. Missing returns occur on revising dates or as a result of mergers, reversed splits or other corporate actions.
A.3 OSEAX
The Oslo Stock Exchange All-Share index consists of all listed shares on the stock exchange. The index is adjusted for dividend payments and other corporate actions. The OSEAX includes all stocks on Oslo Stock Exchange and is comparable to the union of sets A, B, and C.
Appendix B: Alternative variable for mutual fund trading
In this Appendix, we use an alternative variable for trading made by mutual funds. If signings and redemptions are equally large during a month, our preferred variable in the paper will show no flow-induced trading by mutual funds. In this Appendix, we define the variable as the sum of signings and redemptions. The sum of the signings and redemptions is calculated as
$$\hat{f}_{j} = \sum\limits_{i=1}^{N}(inflows_{t,i}+outflows_{t,i}), j =index,mutual, $$
where both inflows and outflows take non-negative values. With this alternative variable specification, we estimate
$$ \sigma_{i,t}^{2} = \beta_{0} +\beta_{1}\hat{f}_{index,t}+\beta_{2}\hat{f}_{mutual,t}+\beta_{3}v_{ETF,t}+\boldsymbol{\beta_{4}X_{t}}+\epsilon_{t}, i=A,B,C, $$
(5)
where X is a vector of control variables and 𝜖 is the error term. Table 9 shows estimated results for Eq. 5.
Table 9 This table reports estimation results of return variances in response to flows to mutual funds (\(\hat {f}_{index}\) and \(\hat {f}_{mutual}\)) and trading volume in ETFs (v
E
T
F
)
Furthermore, we investigate the relationship between index-linked trading and return variances on portfolio A relative to portfolios B and C by estimating
$$ \sigma_{i,t}^{2} = \beta_{0} +\beta_{1}\hat{f}_{index,t}+\beta_{2}\hat{f}_{mutual,t}+\beta_{3}v_{ETF,t}+\boldsymbol{\beta_{4}X_{t}}+\epsilon_{t}, \quad i=AB,AC,BC. $$
(6)
Table 10 shows estimated results for Eq. 6. Again, we find positive correlation between ETF-trading and return variances for all portfolios.
Table 10 This table reports estimation results of return variances in response to flows to mutual funds (\(\hat {f}_{index}\) and \(\hat {f}_{mutual}\)) and trading volume in ETFs (v
E
T
F
)