---
title: "How do households invest on behalf of their children? Evidence from a robo-advisor"
authors:
  - name: "Alexis Direr"
    affiliation: "Université d'Orléans, LEO"
    orcid: "0000-0002-4459-7780"
  - name: "Indigo Jones"
    affiliation: "Université d'Orléans, LEO"
date: "2025-01-22"
doi: "10.2307/48857613"
keywords: [delegated portfolio choice, son preference, gendered risk preference, robo-advisor, household finance, intergenerational transmission, parental investment]
language: en
type: research-article
---

# How do households invest on behalf of their children? Evidence from a robo-advisor

**Authors**
- Alexis Direr — Université d'Orléans, LEO — *alexis.direr@univ-orleans.fr* — ORCID: [0000-0002-4459-7780](https://orcid.org/0000-0002-4459-7780)
- Indigo Jones — Université d'Orléans, LEO — *indigo.jones@univ-orleans.fr*

**DOI**: [10.2307/48857613](https://doi.org/10.2307/48857613)

**Keywords**: delegated portfolio choice, son preference, gendered risk preference, robo-advisor, household finance.

**Research framework**: PREF research initiative under the aegis of the Europlace Institute of Finance, in partnership with Yomoni.

---

## Abstract

Despite the practice being commonplace, little is known about how parents invest in financial markets on behalf of their children. Using a large dataset from the leading French robo-advisor, we find that fathers are more likely to open investment accounts for their sons than their daughters for those aged 12 and above. Since fathers predominantly manage children's contracts within the family, this results in a higher number of savings contracts opened for boys. Additionally, although fathers tend to choose riskier investment profiles for their children compared to mothers, no discernible difference in investment strategy is observed between sons and daughters for either parent.

---

## 1. Motivation and contributions

Investing in financial markets on behalf of one's child is encouraged by many public policies (529 Plan in the U.S., Junior SIPP in the U.K., Livret Jeune in France) and structured legally by the distinction between assets owned by parents and assets owned by the child but custodied by parents until age of majority. Yet the household finance literature has paid little attention to how parents manage these accounts.

The paper exploits extensive data from the leading French robo-advisor and makes five contributions.

1. **Measuring the persistence of gender bias in wealth transmission.** Western legal systems, including the French Civil Code (réserve héréditaire), now treat sons and daughters equally. Cultural norms have likewise shifted. Yet the paper documents a persistence of differential within-family treatment in wealth management and transmission.

2. **Testing a principal-agent problem mitigated by altruism.** Choosing a risk profile on behalf of a child is a principal-agent relationship in which moral hazard and incentive misalignment are presumably minimized (parental altruism, parents' own money). Still, experimental evidence shows that people are weakly more risk-averse when dealing with others' money (Eriksen et al. 2020; Eriksen & Kvaløy 2010). Direr & Visser (2016) find that male financial advisors steer their clients toward riskier assets than female advisors do. The paper transposes this result to the parental relationship: **fathers choose riskier portfolios than mothers for their children** — meaning the same child can end up with a different risk exposure depending on which parent opens the account.

3. **Documenting intergenerational transmission of financial values.** A sizeable literature establishes correlations between parents' and children's risk preferences, saving behaviors and asset ownership (Dohmen et al. 2011; Charles & Hurst 2003; Kimball, Sahm & Shapiro 2009; Arrondel 2013; Fagereng, Mogstad & Rønning 2021). Opening a contract on behalf of one's child can establish lifelong saving habits, signal trust in financial markets, and provide first-hand financial experience (Jorgensen & Savla 2010; Webley & Nyhus 2006). Our results indicate that **fathers favor their sons** in this financial transmission.

4. **Identifying early gender differences in risk-taking.** Traditional norms tend to encourage boys toward more adventurous behavior and girls toward greater caution (Caswell, McKinney & McLaren 2015; Rosvall et al. 2018), with potential consequences for adult financial confidence (Barber & Odean 2001). Contrary to this hypothesis, the paper finds **no** difference in risk profiles between sons and daughters within the same parent's choices.

5. **Identifying who in the household manages children's savings.** Women on average spend more time on childcare, while men make the majority of household investment decisions. Investing on behalf of one's child sits at the intersection of these two roles, so allocation of responsibility is not obvious *a priori*. The paper shows that **men disproportionately manage child investment accounts**: fathers open about 4 out of 5 child contracts.

---

## 2. Data and subscription process

**Source.** Leading robo-advisor in France (Yomoni). Data: August 2015 – April 2022.

**Samples used in the gendered analyses (Section 4):**
- **2,792** parents who opened at least one child account (parent-child investors).
- **3,033** child accounts linked to at least one parent account (parent-child investors' children).

The full descriptive sample (Table 2) covers five groups: non-parent investors (N=20,415), parent-only investors (N=9,895), parent-child investors with at least one child account (N=2,792), their children (N=3,033), and child-only accounts without identified parents (N=1,748).

**Identity verification.** Opening a child account requires an up-to-date *livret de famille* and the child's ID, so the parent-child relationship and demographic characteristics are **externally verified** — unlike financial information, which remains self-reported.

**Process.** The client first selects an investment project among: *Increase savings, Prepare a major purchase, Bequeath an inheritance, Plan their retirement, Save in the event of hard times, Prepare a real estate investment, Finance their children's studies, Open an account for one's child*. If the latter is selected, the savings belong to the child, who will have full access at age of majority; questions about the investment concern the child, while questions about market knowledge, risk appetite and liquidity needs concern the legal representatives.

### Asset allocation by risk profile (Table 1)

| Profile | Money market | Bond ETFs | Stock ETFs |
|:---:|:---:|:---:|:---:|
| 1 | 100% | 0% | 0% |
| 2 | 70% | 15% | 15% |
| 3 | 60% | 20% | 20% |
| 4 | 40% | 30% | 30% |
| 5 | 20% | 40% | 40% |
| 6 | 0% | 50% | 50% |
| 7 | 0% | 40% | 60% |
| 8 | 0% | 30% | 70% |
| 9 | 0% | 20% | 80% |
| 10 | 0% | 0% | 100% |

**Update notes.** Profile 1 is no longer selectable since September 10, 2020 (existing contracts were preserved).

---

## 3. Descriptive statistics on parent and child accounts

Parents who open a contract on behalf of a child differ systematically from other adults:

- **More often men** than other subscribers (84% vs. 70–72%).
- **Wealthier**: higher income, homeownership, financial wealth and property assets than non-parents and child-only accounts.
- **Higher risk profiles**: 7.42 on average for parents with a child account, vs. 7.07 for parent-only investors and 6.88 for non-parents.
- **Higher financial knowledge** on the three quiz questions.
- **Even higher risk profiles for child accounts** (7.60 on average), partly explained by longer investment horizons (15 years on average vs. 11 for non-parents).
- **Much smaller initial deposits** for child accounts (≈ €2,400) than for adult accounts (≈ €8,000–€11,000).

These differences justify a targeted within-family analysis rather than a direct comparison with the general adult subscriber population.

---

## 4. Stylized facts on gender differences

### 4.1 Subscriptions

The analysis focuses on families with at least one identified adult account and one child account.

**Who opens the account?** Fathers open the vast majority of child accounts:
- **87.5%** for children aged 0–5
- **80.7%** for ages 6–11
- **79.6%** for ages 12–17

This proportion is consistent with the share of men among all adult subscribers (at least 2/3, up to 89.7% for the 18–29 bracket, declining with age).

**For whom?** The share of contracts opened for sons is:
- **51.7%** for ages 0–5 (close to parity)
- **49.4%** for ages 6–11 (close to parity)
- **55.1%** for ages 12–17 (visible son preference)

**Decomposition by parent.** Looking only at fathers:
- 0–5: 87.1% of accounts opened for sons vs. 87.9% for daughters (no difference)
- 6–11: 81.9% vs. 79.5% (small difference)
- **12–17: 83.0% vs. 75.3%** — clear son preference among fathers

The mirror reading is that **mothers express a daughter preference for teenagers**. But because fathers subscribe the larger share of child contracts, son preference emerges at the family level.

### 4.2 Risk profiles

**No son/daughter gap at the aggregate level:**
- 0–5: 7.77 (boys) vs. 7.81 (girls)
- 6–11: 7.95 vs. 7.89
- 12–17: 7.63 vs. 7.46 (small gap)

**But a persistent father/mother gap** — fathers choose riskier profiles than mothers for their children, and the gap **widens with the child's age**:
- 0–5: 7.81 (father) vs. 7.63 (mother)
- 6–11: 7.99 vs. 7.62
- 12–17: 7.67 vs. 7.12

**Full decomposition (father/mother × son/daughter):** neither fathers nor mothers significantly differentiate between sons and daughters. Some marginal gaps remain (mothers: 7.54 boys vs. 7.31 girls at 0–5; fathers: 7.75 vs. 7.55 at 12–17), with no systematic direction.

---

## 5. Econometrics

### 5.1 Logit model of subscriptions (Table 3)

$$
P(SON_i = 1) = f\big(\alpha + \beta_0 DAD_i + \beta_1 AGE1_i + \beta_2 AGE2_i + \beta_3 DAD_i \times AGE1_i + \beta_4 DAD_i \times AGE2_i + \gamma X_i\big)
$$

with $SON_i = 1$ if the contract is opened for a boy, $DAD_i = 1$ if the subscriber is the father, $AGE1_i = 1$ for a child aged 6–11, $AGE2_i = 1$ for ages 12–17.

Three specifications: 1.1 (baseline), 1.2 (+ family size, parent's age, subscription years), 1.3 (+ parent's income and wealth).

**Key result.** Only the interaction term **DAD × AGE_12-17** is significant at the 5% level, with a stable coefficient around **+13.9 to +14.4 points**. Family size, parent's age, income and wealth do not predict the probability of opening a contract for a son. Son preference is therefore specifically driven by fathers, and exclusively for teenage children.

### 5.2 OLS model of risk profiles (Table 4)

$$
RP_i = \alpha + \beta_0 SON_i + \beta_1 DAD_i + \beta_2 AGE1_i + \beta_3 AGE2_i + \beta_4 DAD_i \times AGE1_i + \beta_5 DAD_i \times AGE2_i + \gamma X_i
$$

Four specifications: 2.1 (baseline), 2.2 (+ demographics + years), 2.3 (+ income/wealth), 2.4 (+ parent's risk profile, horizons, liquidity, risk aversion, financial knowledge).

**Key results:**
- **`SON` is never significant** in any specification → no son/daughter difference in chosen profiles, once other variables are controlled for.
- **`DAD` is significant in 2.1 and 2.2** (≈ +0.27 to +0.28 profile points), but **disappears in 2.3 and 2.4** once income/wealth and then risk-aversion and knowledge variables are included.
- **`AGE_12-17` is significant in 2.1** (≈ −0.35), then becomes marginal or non-significant in the full specifications.

**Interpretation.** The raw "fathers take more risk" effect is explained by their characteristics (income, wealth, risk aversion, financial knowledge) rather than by gender *per se*. Once these are controlled for, the parent's gender has no residual causal effect on the chosen profile.

**Significant variables in the full specification (2.4):** parent's own risk profile (+0.08 per point), income, property wealth, risk aversion (strongly negative at high levels: Risk Q2 "Very high": −4.58), past experience of financial losses, response to market dips (Buy: +0.32; Sell: −0.43), liquidity needs.

### 5.3 The role of the algorithm

**69.6% of child accounts follow the algorithmic recommendation** at subscription. Since the algorithm is **gender-neutral**, this high compliance rate largely accounts for the absence of son/daughter gaps in chosen profiles. The result illustrates a **benefit of algorithmic choice**: it mechanically reduces gender bias in financial decisions, provided the upstream questionnaire does not embed such biases.

---

## 6. Conclusion

The paper establishes three main results.

**First — son preference for teenagers.** For children aged 12–17, fathers more often open a contract for their son than their daughter. As fathers manage the large majority (≈ 4/5) of child accounts, this bias translates at the family level into a higher number of contracts opened for boys.

**Second — no profile difference between sons and daughters.** Neither parent systematically differentiates the risk taken for a son or a daughter, once other determinants are controlled for.

**Third — the "fathers riskier than mothers" effect disappears with controls.** Fathers on average choose riskier profiles than mothers, but this gap fades once income, wealth, risk aversion and knowledge variables are included. The key role of the gender-neutral algorithm — followed by 70% of clients — helps mitigate gender bias in portfolio choice.

### Possible extensions

- **Longitudinal study** of children's portfolios beyond subscription: son preference might also appear as **higher contributions** over time, rather than solely through initial account opening.
- **Other financial parent-child channels:** pocket money, opening of bank/saving accounts, financing of higher education, inter-vivos transfers.

---

## Acknowledgments

The authors thank for useful comments Béatrice Boulu-Reshef, Luc Arrondel, Claire Rimbaud, Carolin Hoeltken, and participants of the LEO weekly seminar, the 2022 IdR PREF Workshop on Behavioral Portfolio Choice, the 2023 IdR PREF Workshop on Financial Advice and Risk Preferences, and the 2024 GDRE Money, Banking and Finance Conference. This research has been conducted within the PREF research initiative under the aegis of the Europlace Institute of Finance, a joint initiative by Yomoni.

---

## Main references

Agarwal, S., Muhanna, W., Sanders, A. (2011). Gender differences in risk preferences, financial literacy and portfolio choice. *Journal of Banking and Finance* 35(8), 2159–2175.
Arrondel, L. (2013). Are "daddy's boys" just as rich as daddy? The transmission of values between generations. *Journal of Economic Inequality* 11(4), 439–471.
Barber, B. M., Odean, T. (2001). Boys will be boys: gender, overconfidence, and common stock investment. *Quarterly Journal of Economics* 116(1), 261–292.
Boulu-Reshef, B., Direr, A., von Wilczur, N. (2022). Algorithmic vs. human portfolio choice. *LEO working paper*.
Charles, K. K., Hurst, E. (2003). The correlation of wealth across generations. *Journal of Political Economy* 111(6), 1155–1182.
Charness, G., Gneezy, U. (2012). Strong evidence for gender differences in risk taking. *Journal of Economic Behavior and Organization* 83(1), 50–58.
Croson, R., Gneezy, U. (2009). Gender differences in preferences. *Journal of Economic Literature* 47(2), 448–474.
Direr, A., Visser, M. (2016). Portfolio choice and financial advice. *Finance* 34(2), 35–64.
Dohmen, T., Falk, A., Huffman, D., Sunde, U., Schupp, J., Wagner, G. G. (2011). Individual risk attitudes: measurement, determinants, and behavioral consequences. *Journal of the European Economic Association* 9(3), 522–550.
Eriksen, K., Kvaløy, O., Luzuriaga, M. (2020). Risk-taking on behalf of others. *Journal of Behavioral and Experimental Finance* 26.
Fagereng, A., Mogstad, M., Rønning, M. (2021). Why do wealthy parents have wealthy children? *Journal of Political Economy* 129(3), 703–756.
Jorgensen, B. L., Savla, J. (2010). Financial literacy of young adults: the importance of parental socialization. *Family Relations* 59(4), 465–478.
Kimball, M. S., Sahm, C. R., Shapiro, M. D. (2009). Risk preferences in the PSID: individual imputations and family covariation. *American Economic Review* 99(2), 363–368.
Love, D. A. (2010). The effects of marital status and children on savings and portfolio choice. *Review of Financial Studies* 23(1), 385–432.
Webley, P., Nyhus, E. K. (2006). Parents' influence on children's future orientation and saving. *Journal of Economic Psychology* 27(1), 140–164.

*The full reference list appears in the PDF.*