Share Class Mapping and Line Selection

Alphabet has Class A shares (GOOGL, one vote each), Class C shares (GOOG, no votes), and Class B shares held mostly by founders and not publicly traded. They're economically almost identical — same claim on the company's earnings — but they are different instruments with different tickers, different liquidity, and sometimes different prices. A model, an index, or a risk system that says "hold Alphabet" has to decide which line it actually means, and that decision is called line selection.

This isn't a cosmetic detail. Index providers publish explicit rules for it: some indices include only the most liquid share class to avoid double-weighting one company, others include multiple classes but cap the combined weight. A backtest that naively pulls "Alphabet" price history from a database and gets Class A one year and Class C the next — because a vendor changed its default mapping — will show a return discontinuity that has nothing to do with the business and everything to do with a silent line switch.

The practical rule most desks and index committees converge on: pick one class as the primary economic proxy (usually the one with the deepest volume and tightest spread) and map every reference to the company onto that line consistently over time, documenting the choice. Where multiple classes are genuinely held, they should be treated as separate line items — separate borrow rates, separate liquidity profiles, separate closing prices — never quietly merged into one "average" price, which is a number that doesn't correspond to anything a trader can actually execute against.