Machine learning (ML) mispricing models are designed to detect hidden nonlinearities that are important in predicting the fundamental value of stocks. As such, they have the potential to outperform corresponding linear regression (LR) mispricing models. Thus, it is important to allow for nonlinearities and interactions in fundamental analysis.
Fundamental analysis is an approach that is used to determine the intrinsic or fair value of a firm and forms the basis of evaluating whether the company is undervalued or overvalued. Investors can potentially gain from such assessments if they subscribe to the notion that a company’s share price converges to its fair value over the long run: either by buying undervalued stocks or selling overvalued ones.
According to the academic literature, fundamental analysis is typically based on highly stylized valuation approaches such as discounted cashflow models which require inputs such as cashflow forecasts and discount rates. This approach is complicated by the discretion a researcher has over the choice of variables and parameters of the model.
Although these stylized models are extremely popular, explicit cash flow forecasts and discount rates are not necessarily required for fundamental analysis. For instance, an agnostic approach can estimate the fair value of a company as a linear function of its balance sheet, income statement and cashflow statement items.
To this end, a direct approach for estimating fair values is proposed by Bartram and Grinblatt in two academic studies.1 They “take the view of a statistician with little knowledge of finance” and use LR to proxy the “peer-implied fair value” of a firm as a linear function of 21 commonly reported accounting items. They conclude in their findings that their signal reliably predicts future returns in the US and most regions in the world, with the exception of the European market.
In a recent research paper,2 Hanauer, Kononova and Rapp opt for a different approach as they “take the view of a data scientist with little knowledge of finance”. Inspired by the studies of Bartram and Grinblatt, they apply LR and ML methods to estimate the monthly fair values of stocks from 17 European countries for the period January 1993 to December 2019. Then, based on the results, they assess the return predictability of the corresponding mispricing signals, i.e., the difference in model-based fair values and actual market values.
In their analysis, the researchers determined the fundamental values of stocks using six different approaches based on:
More specifically, the researchers used RF and GBRT models given that these can deal with nonlinearities and interactions, handle noisy features well, and do not require subtle tuning as is the case for more complex methods.
The researchers sorted the stocks into five quintile portfolios based on the various mispricing signals. They observed that all the models reflected large negative (positive) mispricing signals for the first (fifth) quintile portfolios. Interestingly, the LASSO and ML signals were considerably smaller than their LR counterparts due to the nonlinearity of their valuation models and ability to better fit the data.
To assess the efficacy of the models, they calculated the value-weighted and industry-adjusted monthly portfolio returns to study the relationship between the mispricing signals and subsequent monthly returns. As depicted in Figure 1, they saw that the ML approaches generated statistically and economically significant industry-adjusted return spreads, benefiting uniformly from both long and short positions. While the LR and LASSO signal spreads were significant, their economic relevance was substantially weaker, with a higher portion of their returns coming from the short leg.
The researchers also verified the results by taking into account four different factor models. In their tests, they noted that the returns of the LR strategy were largely explained by the common factors. Similarly, the alphas for the LR (pooled) signal decreased. By contrast, the ML models delivered similar or even stronger alphas across all factor models. As such, ML methods seem to detect hidden nonlinearities that are important in predicting the fundamental value of stocks.
ML methods are expected to discover additional structure in data due to their ability to spot nonlinear patterns. Consistent with this view, this analysis shows that the portfolio spreads based on ML mispricing signals can earn large and significant alphas, and outperform corresponding LR mispricing models. These findings suggest that it is important to allow for nonlinearities and interactions in fundamental analysis.
1 Bartram, S. M., and Grinblatt, M., April 2018, “Agnostic fundamental analysis works”, Journal of Financial Economics; and Bartram, S. M., and Grinblatt, M., January 2021, “Global market inefficiencies”, Journal of Financial Economics.
2 Hanauer, M. X., Kononova, M., and Rapp, M., April 2022, “Boosting agnostic fundamental analysis: using machine learning to identify mispricing in European stock markets“, Finance Research Letters.
The contents of this document have not been reviewed by the Securities and Futures Commission ("SFC") in Hong Kong. If you are in any doubt about any of the contents of this document, you should obtain independent professional advice. This document has been distributed by Robeco Hong Kong Limited (‘Robeco’). Robeco is regulated by the SFC in Hong Kong.
This document has been prepared on a confidential basis solely for the recipient and is for information purposes only. Any reproduction or distribution of this documentation, in whole or in part, or the disclosure of its contents, without the prior written consent of Robeco, is prohibited. By accepting this documentation, the recipient agrees to the foregoing
This document is intended to provide the reader with information on Robeco’s specific capabilities, but does not constitute a recommendation to buy or sell certain securities or investment products. Investment decisions should only be based on the relevant prospectus and on thorough financial, fiscal and legal advice. Please refer to the relevant offering documents for details including the risk factors before making any investment decisions.
The contents of this document are based upon sources of information believed to be reliable. This document is not intended for distribution to or use by any person or entity in any jurisdiction or country where such distribution or use would be contrary to local law or regulation.
Investment Involves risks. Historical returns are provided for illustrative purposes only and do not necessarily reflect Robeco’s expectations for the future. The value of your investments may fluctuate. Past performance is no indication of current or future performance.
Please read this information carefully.
This website is prepared and issued by Robeco Hong Kong Limited ("Robeco"), which is a corporation licensed by the Securities and Futures Commission in Hong Kong to engage in Type 1 (dealing in securities); Type 4 (advising in securities) and Type 9 (asset management) regulated activities. The Company does not hold client assets and is subject to the licensing condition that it shall seek the SFC’s prior approval before extending services at retail level. This website has not been reviewed by the Securities and Futures Commission or any regulatory authority in Hong Kong.
2. Important risk disclosures
2. Important risk disclosures Robeco Capital Growth Funds (“the Funds”) are distinguished by their respective specific investment policies or any other specific features. Please read carefully for the risks of the Funds:
3. Local legal and sales restrictions
The information contained in the Website is being provided for information purposes.
Neither information nor any opinion expressed on the Website constitutes a solicitation, an offer or a recommendation to buy, sell or dispose of any investment, to engage in any other transaction or to provide any investment advice or service. The information contained in the Website does not constitute investment advice or a recommendation and was prepared without regard to the specific objectives, financial situation or needs of any particular person who may receive it. An investment in a Robeco product should only be made after reading the related legal documents such as management regulations, prospectuses, most recent annual and semi-annual reports, which can be all be obtained free of charge at www.robeco.com/hk/en and at the Robeco Hong Kong office.
4. Use of the Website
The information is based on certain assumptions, information and conditions applicable at a certain time and may be subject to change at any time without notice. Robeco aims to provide accurate, complete and up-to-date information, obtained from sources of information believed to be reliable. Persons accessing the Website are responsible for their choice and use of the information.
5. Investment performance
No assurance can be given that the investment objective of any investment products will be achieved. No representation or promise as to the performance of any investment products or the return on an investment is made. The value of your investments may fluctuate. The value of the assets of Robeco investment products may also fluctuate as a result of the investment policy and/or the developments on the financial markets. Results obtained in the past are no guarantee for the future. Past performance, projection, or forecast included in this Website should not be taken as an indication or guarantee of future performance, and no representation or warranty, express or implied, is made regarding future performance. Fund performance figures are based on the month-end trading prices and are calculated on a total return basis with dividends reinvested. Return figures versus the benchmark show the investment management result before management and/or performance fees; the fund returns are with dividends reinvested and based on net asset values with prices and exchange rates of the valuation moment of the benchmark.
Investments involve risks. Past performance is not a guide to future performance. Potential investors should read the terms and conditions contained in the relevant offering documents and in particular the investment policies and the risk factors before any investment decision is made. Investors should ensure they fully understand the risks associated with the fund and should also consider their own investment objective and risk tolerance level. Investors are reminded that the value and income (if any) from shares of the fund may be volatile and could change substantially within a short period of time, and investors may not get back the amount they have invested in the fund. If in doubt, please seek independent financial and professional advice.
6. Third party websites
This website includes material from third parties or links to websites maintained by third parties some of which is supplied by companies that are not affiliated to Robeco. Following links to any other off-site pages or websites of third parties shall be at the own risk of the person following such link. Robeco has not reviewed any of the websites linked to or referred to by the Website and does not endorse or accept any responsibility for their content nor the products, services or other items offered through them. Robeco shall have no liability for any losses or damages arising from the use of or reliance on the information contained on websites of third parties, including, without limitation, any loss of profit or any other direct or indirect damage. Third party off-site pages or websites are provided for informational purposes only.
7. Limitation of liability
Robeco as well as (possible) other suppliers of information to the Website accept no responsibility for the contents of the Website or the information or recommendations contained herein, which moreover may be changed without notice.
Robeco assumes no responsibility for ensuring, and makes no warranty, that the functioning of the Website will be uninterrupted or error-free. Robeco assumes no responsibility for the consequences of e-mail messages regarding a Robeco (transaction) service, which either cannot be received or sent, are damaged, received or sent incorrectly, or not received or sent on time.
Neither will Robeco be liable for any loss or damage that may result from access to and use of the Website.
8. Intellectual property
All copyrights, patents, intellectual and other property, and licenses regarding the information on the Website are held and obtained by Robeco. These rights will not be passed to persons accessing this information.
10. Applicable law
The Website shall be governed by and construed in accordance with the laws of Hong Kong. All disputes arising out of or in connection with the Website shall be submitted to the exclusive jurisdiction of the courts of Hong Kong.