Amid a spate of high-profile crypto project failures, skeptics are asking what crypto is good for other than speculative gambling. Although some see the onset of “crypto winter” as vindicating suspicions that the space was always just one big casino, the very issues highlighted by recent crypto firm bankruptcies – questionable risk management, opaque balance sheets, and crypto deposits’ vulnerability in bankruptcy – actually help to answer the question: what’s the point of crypto?

In a word, decentralization. While this term can be employed in handwavy hype, the recent bear cycle helped to reify the benefits of disintermediated finance. Policymakers should pay close attention to the resiliency features of decentralized projects – including automated collateral liquidations, public transaction data, and self-custodied assets – before baking faulty assumptions about crypto risks into regulations.

Decentralized finance or “DeFi” uses cryptocurrencies’ fundamental innovation – the ability to validate peer-to-peer transactions without a trusted third party – to provide financial services without intermediaries. Cryptocurrencies achieve this by replacing institutions like banks with a network of computers incentivized to maintain a transaction history on a cryptographically secure distributed ledger, a blockchain. This technology also supports “smart contracts,” transactions that self-execute when specified conditions are met. Cryptographer Nick Szabo famously compared smart contracts to vending machines: if $1.50 goes in, a soda comes out.

DeFi protocols operating in the wild have applied the vending machine model to finance, offering to users: when deposits go in, interest accrues, and when collateral goes in, loans come out. Unlike traditional finance, DeFi is permissionless – users just need the right cryptocurrency for deposits or collateral, not a credit history or preexisting relationship. Moreover, DeFi doesn’t depend on individuals keeping promises, but rather the execution of code. Arguably, DeFi is crypto’s long-sought “killer app.”

The headline-grabbing bankrupt crypto ventures, hedge fund Three Arrows Capital (3AC), and lenders Voyager Digital and Celsius Network were not DeFi projects. Although they traded in DeFi tokens or borrowed from DeFi protocols, each was a traditional firm, not a series of smart contracts.

Their bankruptcies should sound familiar to financial historians. Lenders Voyager and Celsius facilitated 3AC’s leveraged crypto asset trading, including through an unsecured loan from Voyager. When asset prices plummeted, 3AC didn’t meet margin calls and defaulted on loans. In the downturn, the lenders faced depositor withdrawals which, when combined with 3AC’s nonpayment, contributed to insolvency.

A novel aspect of the crypto lenders’ bankruptcies is the impact on depositors, who may face challenges recovering their assets. Although the lenders functionally resembled, to varying extents, banks and brokerages, they were not squarely covered by special banking or trading protections for customer assets during bankruptcy.

To some, this harsh legal reality demonstrates the excessive risk of crypto finance. But using centralized platforms is a choice, not a necessity. Cryptocurrency users can avoid intermediaries by directly accessing, with a unique private key, individualized blockchain accounts. Unlike storing cash under a mattress, this “self-custody” option keeps funds portable – they remain accessible around the world to the individual with the correct private key.

Another downside of keeping cash under a mattress is the inability to earn interest, and centralized crypto lenders pitched earning yield as one of their offerings. But DeFi protocols can compensate users for contributing to liquidity pools that facilitate decentralized lending and asset exchange. Notably, depositing through smart contracts carries risks, like potential hacking or losses from asset price swings.

Nonetheless, while centralized crypto firms’ bankruptcies garnered attention, notable DeFi projects kept working through the recent bear cycle. A combination of features may have contributed to this resilience, including that DeFi protocols typically are incentivized to require overcollateralized loans, for which debtors pledge more value in one cryptocurrency than they borrow in another, and automated liquidations, in which liquidators, according to smart contracts, help repossess collateral and repay creditors if the collateral loses too much value relative to the loan. In addition, while centralized lenders’ risk-laden balance sheets can blindside depositors, smart contracts are public by design.

One can’t know for certain how any specific project, let alone an entire ecosystem, will fare long term. The point is not to spotlight particular DeFi projects and blast bankrupt centralized firms, as there are failed DeFis and successful centralized crypto projects. Nor is the point to argue that DeFi is always preferable to centralized finance – that almost certainly will vary by users’ needs.

Rather, the point is to understand DeFi’s unique capabilities, so that when policymakers look to address financial risks, they understand the differences between centralized firms and DeFi projects. This means not only understanding DeFi’s risks – including complex (if public) transaction histories, cybersecurity vulnerabilities, and inefficiencies and volatility risks in its approach to collateral – but also understanding decentralization’s capacity to counteract other risks by opening up transaction data, incentivizing conservative approaches to secured lending, and allowing individuals to self-custody digital assets. DeFi may not be right for everyone, but crypto is revolutionary for making finance without middlemen an option. When it comes to financial innovation, regulators should consider choice a feature, not a bug.