While the nominal return on bank deposits has historically been positive or at least zero, two long-run developments suggest that households may need to pay banks to hold their deposits in the future. First, monetary policy rates have fallen steadily since the early 1980s and have recently been in negative territory for sustained periods in some of the world’s largest economies (Lilley and Rogoff 2020). Second, the digitalisation of the financial system has rendered bills and coins an unpractical means of payment and wealth storage. While negative monetary policy rates put pressure on banks to pass their costs of holding liquidity on to depositors (Krogstrup et al. 2020), the rise of the digital economy removes an important practical barrier to setting negative deposit rates.

How should we expect households to behave in an environment with deposit rates below zero? In standard models of consumption and saving, households respond to real rather than nominal returns and there is nothing inherently special about zero: a change from 0% to -1% is conceptually the same as a change from 2% to 1% or from 1% to 0%. However, access to a zero-return asset in the form of cash and behavioural frictions such as nominal loss aversion and a perceived unfairness of banks charging households to hold their deposits suggest that responses to negative interest rates may be different. 

The question is important for monetary transmission, but the literature is scant. While several papers study the behaviour of banks in environments with negative policy rates (Ulate 2021, Abadi et al. 2023, Eggertsson et al. 2024), only few papers analyse how firms respond when banks pass negative rates on to deposits (Altavilla et al. 2022, Abildgren and Kuchler 2023), and there is almost no work on households.

Negative deposit rates in Denmark

In a recent paper, we study household behaviour in the context of a major episode of negative interest rates in Denmark (Andersen et al. 2026). In 2019, the first Danish bank introduced negative rates on personal deposits, applying -0.75% to balances exceeding DKK 750,000 (around $115,000). In the following years, virtually all banks adopted negative rates, and balance thresholds generally declined to DKK 100,000 (around $15,000). As illustrated by Figure 1, around one-third of all individuals and two-thirds of all personal deposits in the banking system were subject to negative rates by the end of 2021.

Figure 1 Exposure to negative deposit rates in Denmark

How do households respond to negative deposit rates?

We first analyse how households adjusted their financial behaviour when they became exposed to negative deposit rates. We use comprehensive administrative micro data, including account-level information covering the universe of bank accounts in Danish banks, and exploit cross-bank variation in interest rate policies for empirical identification. In the main analysis, we compare a treatment group that became exposed to negative rates in 2019 to a control group that became exposed in 2021.

As shown in Figure 2, we find that exposure to negative deposit rates caused a strong decrease in deposits: the treatment group exposed in 2019 differentially reduced their total deposit balances in the banking system by DKK 90,000 during 2019–2020 relative to a baseline of DKK 1.1 million (Figure 2A). This estimate is remarkably large compared to what we find in a parallel analysis, where we estimate depositor responses to rate cuts in positive territory.

Figure 2 Household responses to negative deposit rates

What happened to the funds that flowed out of deposit accounts in response to negative rates? We estimate that around one-third was reallocated to other asset classes: individuals in the treatment group differentially increased their stock market holdings (Figure 2B) and contributions to illiquid pension accounts during 2019–2020 (Figure 2C). However, most of the reduction in deposit balances appears to reflect increased consumption. Imputing consumption from detailed administrative data on income and wealth (Jensen and Johannesen 2017, Holm et al. 2021), we find that the treatment group differentially increased annual consumption by around DKK 40,000 during 2019–2020 (Panel 2D). These estimates suggest a sizeable stimulating effect of negative monetary policy rates – in our case a contribution to aggregate private consumption of as much as 0.5%–1% in 2020 – at the time they pass through to households.

Lacking micro data on cash holdings, we use data on aggregate cash withdrawals to investigate whether households responded to negative deposit rates by holding more cash, a key mechanism in standard theory. Figure 3 shows that cash withdrawals were generally at, or slightly below, their long-run trend during the period with negative rates, suggesting that the liquidation of deposits to hold cash was not a quantitatively important response. This is consistent with non-trivial costs of holding cash in large amounts (Rogoff 2017).

Figure 3 Aggregate cash withdrawals at the introduction of negative deposit rates

How do households reason about deposit rates?

To discriminate between the various motives that may underlie the depositor responses – for example, standard motives such as intertemporal substitution versus non-standard motives such as loss aversion or perceived unfairness of negative rates – we conduct a survey. We ask respondents whether and how they would adjust different margins of behaviour in a hypothetical scenario where the deposit rate dropped. In open-ended responses, we also ask them to explain in their own words why they would or would not adjust their behaviour as indicated (Haaland et al. 2025). The main scenario mirrors the situation in Denmark in 2021 when all banks adopted negative deposit rates, but we also randomise some respondents into scenarios with rate decreases in the positive domain, which allows us to study systematic differences in responses and motives between rate cuts in negative versus positive territory.

To validate the survey approach, we link survey responses and administrative data and show that the respondents’ self-reported intentions to make adjustments in response to a hypothetical rate cut predict actual changes observed in the administrative data at the time they became exposed to negative rates in the real world. Moreover, consistent with the results from the analysis of administrative data, the survey responses indicate that rate cuts in the negative domain are much more likely to trigger adjustments than comparable rate cuts in the positive domain.

Figure 4 illustrates the open-ended responses on the motives underlying consumption responses, grouped into broad classes. Most respondents who indicate they would increase consumption in response to negative rates cite a distaste for losses (Gneezy 2005), anger about perceived greed or unfair behaviour by banks, or more standard intertemporal substitution arguments. Those who would not change their consumption often cite a general insensitivity to interest rates, for example due to adjustment frictions or because they are satisfied with their current level of consumption.

Figure 4 Motives behind consumption responses to negative rates

Similarly, Figure 5 shows the open-ended responses about investment responses. Most respondents who would increase stock investment in response to negative deposit rates cite loss aversion or return differentials. Those who would not adjust stock holdings typically emphasise a general reluctance to participate in the stock market, for example due to participation costs, lack of knowledge, or the perceived riskiness of stocks.

Figure 5 Motives behind stock market responses to negative rates

Importantly, the two non-standard motives behind adjustments to negative rates – loss aversion and perceived unfairness – are almost absent in the scenarios featuring rate cuts in non-negative territory. This suggests that households think fundamentally differently about nominal rate cuts in negative territory.

Implications for monetary policy transmission

These findings have important implications for monetary policy transmission in low-rate environments. Recent theory contributions assume that negative policy rates do not pass through to deposit rates, since households would switch to cash if they did. According to these models, the consumption effects of rate cuts in negative territory are therefore muted (e.g. Ulate 2021) or possibly even reversed (Abadi et al. 2023, Eggertsson et al. 2024). Our results only partly agree with this view. 

On the one hand, we find support for the notion that zero is a critical bound, below which depositors respond more strongly to rate changes, potentially disciplining banks’ decisions about pass-through. On the other hand, the nature of depositor responses is to consume more rather than to hold more cash, suggesting that, conditional on pass-through, policy rate cuts in negative territory may have particularly strong consumption effects.

References

Abadi, J, M Brunnermeier, and Y Koby (2023), “The reversal interest rate”, American Economic Review 113(8): 2084–120.

Abildgren, K, and A Kuchler (2023), “Firm behaviour under negative deposit rates”, European Economic Review 151: 104349.

Altavilla, C, L Burlon, M Giannetti, and S Holton (2022), “Is there a zero lower bound? The effects of negative policy rates on banks and firms”, Journal of Financial Economics 144(3): 885–907.

Andersen, A L, N Johannesen, J B Petersen, S Settele, and J Wohlfart (2026), “Household behavior below the zero lower bound”, CEPR Working Paper 21623.

Eggertsson, G B, R E Juelsrud, L H Summers, and E G Wold (2024), “Negative nominal interest rates and the bank lending channel”, Review of Economic Studies 91(4): 2201–75.

Gneezy, U (2005), “Deception: The role of consequences”, American Economic Review 95(1): 384–94.

Haaland, I, C Roth, S Stantcheva, and J Wohlfart (2025), “Understanding economic behavior using open-ended survey data”, Journal of Economic Literature 63(4): 1244–80.

Holm, M B, P Paul, and A Tischbirek (2021), “The transmission of monetary policy under the microscope”, Journal of Political Economy 129(10): 2861–904.

Jensen, T L, and N Johannesen (2017), “The consumption effects of the 2007–2008 financial crisis: Evidence from households in Denmark”, American Economic Review 107(11): 3386–414.

Krogstrup, S, A Kuchler, and M Spange (2020), “Negative interest rates: The Danish experience”, VoxEU.org, 2 October.

Lilley, A, and K Rogoff (2020), “Negative interest rate policy in the post COVID-19 world”, VoxEU.org, 17 April.

Rogoff, K (2017), “Dealing with monetary paralysis at the zero bound”, Journal of Economic Perspectives 31(3): 47–66.

Ulate, M (2021), “Going negative at the zero lower bound: The effects of negative nominal interest rates”, American Economic Review 111(1): 1–40.