How to choose a validator
Choose the best validator to delegate according to your own profile (risk/reward ratio, liquidity, effort required)
AVAX delegation is good business, up to 12% per year. But choosing a validator is a very delicate choice: you’re putting your future rewards in someone else’s hands. Despite appearances, Avalanche delegation is not 100% trustless: if your validator’s uptime drops <80% (for technical issues or - worse - deliberately because it’s an attacker) you’ll lose all your rewards, and there’s no option to unstake before the staking period ends.
Staking rewards depend on:
- Duration: the longer duration, the higher reward
- Uptime: must be > 80%
Choice depends on:
- Risk profile
- Stake amount
- Effort profile
- Liquidity profile
The goal of this post is to help you choose the best validator to delegate according to your own profile (risk/reward ratio, liquidity, effort required).
Validators' maximum weight is set at 3M AVAX and a validator can’t collect delegations higher than 5 times its stake. You therefore need to make sure your validator has enough capacity to accept your delegation.
Your delegation strategy is mainly affected by your sale option and subsequent lockup period. Duration is one of the most important aspects of delegation. It affects risk, reward, liquidity and effort.
Higher staking period means higher rewards (up to a year). But since the tokens locked cannot be unlocked until the staking ends, you will need to think which period is the best for you. Bear in mind that it’s actually better to sync the staking period with the vesting period of the tokens you purchased in the token sale. Since the vesting period is 1 or 1.5 years, it may be better for you to stake for 1 year and then renew the stake with the new vested tokens.
Another important factor is how your amount staked relates with the tx fees you’re going to pay when renewing your delegation.
- Short term: less risk, more liquidity / more effort, less reward, more tx fees
- Long term: more risk, less liquidity / less effort, more reward, less tx fees
Depending on your sale allocation, there are different delegation durations that are optimal.
On the basis of your allocation:
- Public sale A1: you’re locked for 1 year, you should delegate long term (e.g. 1 year) to optimize rewards;
- Public sale A2: you are locked for 1.5 years, you should delegate long term (e.g. 1 year) to optimize rewards;
- Public sale B: you are fully liquid. You should choose between delegating long term (e.g. 1 year) to earn high rewards, or short term (e.g. 1 day) earning small rewards but preserving liquidity.1. RISKS
Despite appearances, Avalanche delegation is not 100% trustless: if your validator’s uptime drops <80% (for technical issues or - worse - deliberately because it’s an attacker) you’ll lose all of your rewards and you won’t be able to unlock your AVAX until delegation ends.
Imagine Mallory (fake name), a nasty Avalanche competitor, wants to damage Avalanche reputation. She buys 200,000 AVAX OTC, spawns dozens of low fee (0%, 0.1%, 1% etc) validators with long duration (10, 11, 12 months), collects lots of delegations from unwitting delegators. They happily proceed with the delegation, forget about it and plan to come back after a few months. When she collects 50M AVAX in delegations she turns her many nodes off, all at once. Since Avalanche has no slashing function, she won’t lose her 200,000 AVAX. Unfortunately all of her delegators will not only lose all rewards, but their AVAX will be locked until delegation ends. Avalanche will suffer a reputational loss, while delegators will lose money and liquidity.
- Is my delegator’s fee 0%?
- Can I double check the name and website of my validator?
- What’s their historical track record?
- Is my validator’s fee very low? If so, is it enough to pay for ongoing technical expenses and additional services?
- Is my validator’s staking duration higher than 6 months?
You need to understand how likely it is for your validator to go offline for technical faults, its contingency plans, and how large and skilled the team is.
- What’s the specific hardware setup of my validator?
- Is the hardware powerful enough to handle Avalanche peak traffic spikes (up to thousands of transactions per second)?
- Where is the node running and how secure is the location?
- Is the infrastructure redundant?
- What’s its backup policy?
- Is the node protected from DDoS?
- How is the node protecting its private key?
- What happens if a validator operator can’t, for any reason (illness, arrest, death, etc), access and maintain his node?
- Is the manager of the node able to maintain the validator, upgrade it and fix is quickly?
- Financial risk:
- What happens if your validator doesn’t have enough money to keep maintaining increasing hardware costs?
- Regulatory risk:
- What jurisdiction is my node operator in?
- Is the node physically under the same jurisdiction?
- What happens if law enforcement, hosting provider, ISP chooses to stop your validator?
There are endless risks, as with any endeavor. Don’t fall for the illusion there are no risks.
Do the math of your validator: Check the total delegated and the fee. How much will your validator earn per month? Do they offer additional services or one-on-one support?
If that’s not enough to maintain a professional service level, ask questions.
Be suspicious of new nodes (that have no track record), with unknown identity (no name, no website, etc.)
- Am I sure this validator is legit?
- Are they anonymous?
- Is their website trustworthy?
- Do they have experience with staking?
- Are they rated or audited?
- Have they been endorsed?
Take a look at the rewards’ track record: it’s the most important aspect. Look at the last few months. That is not an indicator of future rewards, but a good metric nonetheless.
If the validator has a flawless track record for months or years, they are less likely to go offline unless they’re malicious. On the other hand, if they’re a malicious attacker, they will deliberately show a clean track record until they attack.
- How long has my validator been active on Avalanche?
- What’s its historical uptime for the last 12 months (not its current one)?
- What’s its historical uptime for the last 6 months (not its current one)?
- What’s its historical uptime for the last 3 months (not its current one)?
The other important aspect is the uptime track record: look at the last few weeks / months. Not an indicator of the future, but a good metric here as well. And check if the validator is satisfying the uptime requirements of the network. On mainnet, there’s a threshold of 60% uptime. If a validator stays below 60%, they won’t get any staking rewards, and neither will you as its delegator.
Then you need to know more about subnets: how many subnets are they validating? Are they able and willing to add new subnets in the future?
- Will they share rewards from new subnets with me?
- Will they share fork rewards with me?
- Will my validator offer additional rewards beyond AVAX?
- Will my validator offer additional services?
Avalanche will be governed on-chain, meaning that stakeholders will be able to take key decisions on how it should evolve. When you delegate your AVAX you also delegate your voting power. You validator will use its power to go to consensus about the uptime of other validators, to decide whether they will get the staking rewards - your decision will have an impact on the whole network!
- How will my validator use this power?
- Will I have a voice when key decisions happen, or will the validator choose what’s best for me?
- In the event of an unexpected emergency (imagine The DAO incident and subsequent decisions) will my voice be heard, or my validator’s?
You might need support to delegate, renew your delegation, secure your funds, etc: A validator with an active support is a great plus.