Portfolio Returns, Rebalancing, and Performance Attribution
Summary
The discussion considers how to calculate simple and logarithmic returns for a portfolio, including what changes when holdings are added, removed, replaced, or rebalanced during a measurement period. It emphasizes comparing like with like across the start and end dates, and directs readers to established finance and performance-measurement materials for fuller treatment.
It also highlights complications that can affect reported returns: transaction costs, investor cash flows in funds, and illiquid assets without observable bids. Attribution of portfolio performance to individual assets is identified as a separate discipline. The answer offers guidance rather than derivations or worked examples, and warns that reported historical performance depends on methodology; it points to formal standards as a way to audit performance reporting.
Key ideas
- Portfolio return calculations become more complex when holdings or weights change during the measurement period.
- Rebalancing and asset substitutions require consistent comparisons between beginning and ending portfolio exposures.
- Transaction costs, investor flows, and illiquid valuations can affect measured performance.
- Explaining asset-level contributions to total return is a performance attribution problem.
- Performance reporting choices can alter how historical results appear.
Tags
Full text
# How to compute simple and log portfolio returns? # How to compute simple and log portfolio returns? I am looking for more details to perform simple and log returns for an entire portfolio. However, I've only been able to find the following semi-reliable source (see Page 9 and Page 19): Here are my questions - Are the details for calculating portfolio returns described in the PDF correct? I understand the math but don't trust the source. - Are there better resources that can confirm these calculations? - What do I need to do/alter in the calculation if there is a rebalancing of the portfolio (or a set of assets are removed/added/replaced) in between the start and end dates? - When using portfolio log returns, is there a way to attribute increases/decreases in returns to specific assets within the portfolio (i.e., to say that assets A, B, and C increased by 3%, 4%, and 10%, respectively which resulted in an overall portfolio return of +5%)? ## Answer by SRKX (score 2) https://quant.stackexchange.com/a/21548 Computing returns is one of the first things you learn when you start studying finance but I believe it's one the trickiest one once you get to complicated cases. The source you mentioned seems actually very good to me and it already takes into account different approaches and different subtleties like dividend payment. But this is in fact only the top of the iceberg, because there are many things that make computing returns of portfolios complicated. For example: - Transaction costs which can happened at trade or portfolio level and which can be expressed as a percentage of flat fees. - For funds, having investors buying or selling shares makes it more difficult for you to express your portfolio returns. - You have might be very illiquid assets for which there are no bid (nobody wants to by it), what the price then? - Many more examples I'm sure more experienced community members could be able to highlight. For examples on how to compute portfolio returns on some these more advanced cases, I'd suggest you to look at the CFA Institute and their material. To be fair it might even be possible that the best book for this would not actually be finance book but some sort of accounting book. Finally, note that computing returns is something that somehow depends on interpretation. There are many different ways to compute and present past performance and that has be used many times in the past to make products look better than they really are (or at least, has helped to hide some products' weaknesses). This is why some organization exist now to kind of audit the way performance is reported, such as GIPS. To answer your questions: - The source seems fine to me, although I didn't read it completely so if you have something specific you're worried about you can enhance your question. - Better I'm not sure. But you can have a look at the CFA curriculum books. - You need to take this into account to make sure your comparison "ending price vs initial prices" is apple-to-apple. - This is performance attribution and it's a whole topic on its own.
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.