Russell 2000 consituents

General
1

How come the number of stocks in the Russell 2000 starts at almost precisely 2000 in June each year and then falls to around 1850by hte following May and hten resets again. Is that from stocks being acquired? Why else would they continuously fall out of the index? Or is there a data issue?

This screen / backtest shows the effect.
Screener - Portfolio123

2

This simulation demonstrates the problem more clearly:

Performance - Statistics - Simulated Strategy

It uses the Russell 2000 universe, buys all constituents, and whether you do float cap weighted or mkt cap weighted, there large deviations from the IWM ETF benchmark returns each week. Why isn’t it more precise? Several weeks with +/- 2% deviation from the benchmark return. I’m trying to lower my tracking error and this is part of the problem it seems.

3

The data is proprietary but I think the index provider releases it each june. If you look at the preset universes they say “prussell” not russell. They will not track it perfectly since you are comparing the Russell 2000 to something else.

1 Like
4

A few things make this challenging.

  1. Stocks delisting, takeovers, mergers in between official reconstitution points means the index has declining #’s in months between June
  2. Your sim is not synced up to the reconstitution date
  3. Your sim has fewer holdings that the current P123 Prussell 2000 universe. Not sure why but is this an issue with not enough cash to buy positions as it is fully invested?
  4. An actual sim will have cash drag vs an index which does not

The screener would be better if you wanted to replicate the index. Unfortunately, the screener does not do cap-weighting.

But if you compared (in the screener) holding all positions equal-weight in the S&P 500 and compare this to the RSP (equal-weight S&P 500 ETF) it is similar. But there is still the issue of trying to sync up the rebalance and reconstitution points which are based on quarterly calendar dates as opposed to the 13-week setting in P123.

Small differences will add up.

1 Like