Mongo going out a limb.
tl;dr; TSLA will ultimately be added to each index fund as a fixed number of shares and the share price is a non factor regarding the number of shares added. The share price will impact the final % of the index TSLA comprises by dollar value, but that is a separate detail.
High confidence facts:
- S&P 500 is a float weighted index
- The float market cap is IWF * total shares * stock price
- IWF (Investable Weight Factor) is trading_shares/ total_shares
- IWF is not changed continuously
- Percentage in the Index is based on float_cap/total_index_cap (including itself, which has a small impact on the math).
When dealing with inclusion and % of index, the key thing to remember is that that individual stock holdings are determined purely by math, they are not dictated by the committee. Tesla at 1.5% or 1.0% is based purely on the ratio of TSLA float cap to the entire index's float cap. Each index fund has the IWF, stock price, and share count needed to calculate the individual holdings they should have.
Starting with an easy to answer question:
As the stock price changes, the value percentage of a constituent in the index changes, so how/when do index funds internally adjust to match the target percentage?
Answer: Once included at the proper share level, there is no adjusting needed:
Imagine, for discussion, there is a stock that is 10% of the index:
A: IWF*SP*shares = 10
Everything else = 90
Index total value = 100, A is 10% of total
Now, that stock doubles
A: 20
Everything else: 90
Total fund: 110
Actual percentage of A = 20/110 = 18%
Should the fund sell 8% to get to 10? Buy 2% to get to 20?
Neither! The fund has exactly the position the stock should have. The amount of stock the fund should have is the stock's float adjusted cap/ total index cap. That is the 18.2% number. So the index funds are self adjusting (to stock price changes) without buying/ selling shares.
The difficult question: So what about price changes just before inclusion?
If index funds are told the stock will be 10% of the index based on week old price data, and they convert that to a dollar value, then, if the stock doubles in the interim, they will be 8% off of the correct value.Not good.
However, if they take the % at time of the freeze (which is just the market cap relative to the full index) and convert that to shares based on that index's total value, then, as long as they acquire that number of shares, they end up with the correct percentage of the stock.
This means a rising stock price requires more spending by the funds which follows as the stock has a larger market cap which corresponds to a large index weight. It also means they sell more of the other index constituents to purchase the new stock (total fund value is a constant).
Any staggering they do to achieve the final position is beyond me. Regardless, the ultimate number of shares remains the same and, if the doubled price holds, the percentage of the stock by value will be 18.2% . A spike and pull back impacts the funds profits because more stocks had to be sold to acquire the new one, then the new one lost value, reducing the fund's value. So the # of shares in the new position remains the same, but the end value is less (both for the stock and the index). Same thing as a trading account when you sell one stock to buy another only to have the new purchase drop.
Another way to view this:
Ignore the float weighting for the moment. The index is cap weighted, so the dollar amount of a stock in the index is proportional to its market cap. The total value of all stocks is the index value. So:
value_of_stock_in_index = market_cap_of_stock * value_of_index / total_market_cap_value_of_constituents
value_of_index / total_market_cap_value_of_constituents is a constant (in the short term), let's call it 'a'
value_of_stock_in_index = a * market_cap_of_stock
In other words, the index holds a fixed percentage of each company in the index (and the same percentage of each).
But stock is traded in shares, so how much of each company?
Replace market_cap_of_stock with its formula:
value_of_stock_in_index = a * (total_number_of_shares * share_price)
total_number_of_shares is a constant (unique per stock), call it 'b(x)'
value_of_stock_in_index = a*b(x)* share_price
So the index holds a*b(x) shares of each stock to stay in balance and the share price and value track each other. IWF updates handle when the number of shares change.
My understanding is that you’re MOSTLY correct about the number of shares that need to be purchased being fixed. After they do the weighting it will be fixed, but before then it will not. I’m not sure at what point you meant
In actuality the number of shares decrease slightly with increased market cap relative to the total market cap of all 500 companies.
E.g. Let’s stress test and assume Tesla’s market cap equaled the combined market cap of all the other 499 companies. This would mean that SP500 index funds would now only have enough money to purchase 6.5% of the float of all 500 companies vs. the 13% (my recollection and may not be perfectly accurate) of the float they can currently afford. Or put another way, if Tesla was at $1B per share, they would NOT be buying 100M+ shares of Tesla.
Last edited: