‘Smart’ copycat mutual funds: on the performance of partial imitation strategies

Financial Innovation

Table 6 Imitation by fund quantile

CopyQ	5	4	3	2	1	0	− 1	− 2	− 3	− 4	− 5
5	8.58	5.52	4.67	4.30	4.26	13.17	7.80	8.69	10.02	11.93	21.04
4	6.94	5.16	4.60	4.30	4.50	17.19	8.98	9.39	10.45	11.19	17.30
3	6.05	4.79	4.53	4.38	4.57	20.25	9.38	9.82	10.54	10.79	14.90
2	5.04	4.51	4.45	4.41	4.74	23.62	10.13	10.12	10.48	10.10	12.41
1	4.65	4.22	4.39	4.38	4.83	26.03	10.32	10.20	10.41	9.68	10.89
0	4.06	4.53	4.78	4.95	5.02	37.58	9.44	8.42	8.10	7.18	5.96
− 1	4.85	4.61	4.47	4.62	4.70	29.58	10.43	9.40	9.19	8.65	9.51
− 2	5.55	4.72	4.38	4.48	4.45	27.69	10.22	9.34	9.37	9.14	10.66
− 3	6.65	5.18	4.44	4.41	4.41	24.59	9.79	9.16	9.51	9.80	12.06
− 4	8.26	5.50	4.51	4.52	4.25	20.88	9.17	8.76	9.34	10.44	14.35
− 5	10.45	6.22	4.67	4.57	4.04	16.77	8.13	8.12	9.01	10.95	17.06

Each quarter, all stocks from a fund’s portfolio in which the fund manager is identified as imitating another manager’s trades are classified on two dimensions: the copycat quantile of the fund which holds them, and the quantile of the fund whose trades are imitated. The average group portfolio weight is calculated as the mean of the weight of each stock in its portfolio. The imitating fund’s classification is located on the Y axis, while the quantiles of the funds whose trades they copy are arrayed on the X axis. The table data is normalized so that each row (imitation weights for all funds in a certain quantile) adds up to 100%. The resulting table depicts the intensity with which funds from each quantile tend to imitate funds in every quantile, including their own. For example, of all stocks in which they imitate other funds, managers in quantile 5 tend to devote 8.58% to imitating trades of managers in the same group (5), whereas they follow trades of funds in quantile − 5 21.04% of the time