Skip to content
All library documents

Dot-Com Valuations, Long-Term Returns, and Ex-Ante Bubble Risk

Article Quant Q&A · Author: Kvothe
Read the original

Summary

The document asks whether the original constituents of the Dow Jones Internet Composite Index, held from a date before the dot-com crash, ultimately delivered returns that justified their late-1990s valuations. The proposed comparison would keep the initial constituents fixed, reinvest dividends, and account for delistings and bankruptcies, including possible total losses. No portfolio return calculation or constituent-level evidence is provided.

The answer argues that some internet firms later became highly profitable, but those eventual outcomes do not show that their valuations were justified by information available to investors around 2000. It frames a bubble as excessive expectations and valuations relative to what was knowable at the time, with the subsequent multi-year bear market offered as context. The question draws a parallel with possible AI-stock enthusiasm and asks about the cost of investing before a bubble bursts. The exchange offers a conceptual distinction between hindsight success and ex-ante valuation, but does not quantify long-term returns or establish how to assess current AI valuations.

Key ideas

  • The proposed historical test holds the original index constituents fixed and reinvests dividends.
  • Delistings and bankruptcies need to be included to avoid survivorship bias in a long-term return estimate.
  • Later success by some firms does not prove that their earlier valuations were justified by information available at the time.
  • The answer describes bubbles in terms of excessive expectations relative to knowable evidence.
  • The discussion supplies no return calculation and does not evaluate current AI-stock valuations.

Tags

EquitiesUS marketsStatisticsBacktesting
Full text
# Were internet related companies at the height of the dot-com bubble really overvalued from a long term persepective?


# Were internet related companies at the height of the dot-com bubble really overvalued from a long term persepective?












TL DR: What was the long-term return of the original Dow Jones Internet Composite Index constituents (no rebalancing but reinvesting dividends) from 1999/2000 to today?

This question is motivated by wanting to better understand the dot-com bubble. The expectation around the time was that the internet would be a huge thing resulting in great returns. From the long term point of view this expectation of course turned out to be exactly right. Internet/tech companies indeed showed some of the greatest returns in the times to come. Yet there was a "bubble" that burst and afterwards the perception was that these stocks had been overvalued. There seems to be somewhat of a contradiction in this that I want to understand better. Were these stocks really overvalued (on average/as a whole) when we take the long term perspective or were the valuations at the height of these bubbles actually accurate in the long term?

In turn I want to understand this because the Dot-Com bubble appears very similar to a possible "AI stock bubble". AI is another area where I feel very confident that it will be the most important growth sector in times to come. Yet there might still be some "bubble bursting" before those stocks pay of. I want to better understand how "bad" it is in such a case to invest pre- "bubble burst" from a long-term perspective.

Something that I think would give insight into this question is: What would the long-term total return have been for a portfolio consisting of the original constituents of the Dow Jones Internet Composite Index at a fixed date before the dot-com crash (e.g. Feb 1999, June 1999, or Jan 2000). Some slightly complex things that would need to be considered to come to something that can fairly be compared against the performance of other indices.

- We should probably consider all dividends to be reinvested proportional to the holdings in that year.

- Delisted and bankrupt companies need to be properly taken into account (as total loss?)

Would someone here know how to answer this question?

## Answer by AMach (score 0)

https://quant.stackexchange.com/a/85293

There likely was an actual bubble. Just because a lot of these companies turned out to be massively profitable and established deep moats doesn't mean that was at all likely to have been the case back in 2000. Bubble basically means current valuations and future expectations were way too high given all the knowable information at the time and that became immediately apparent when we entered a multi year bear market following a dramatic crash.

Shown in full with attribution under the source's licence. Licence: CC BY-SA 4.0 (Stack Exchange)

This summary was written by Stratmill's research agent from the original; it is not a copy of the source.