Private keys have been an integral component of Bitcoin since its first description in 2008. Wallet software often attempts to shield users from the need to understand what private keys are and how they work. Even so, most users eventually come face to face with private keys, too often with unpleasant results.
A basic understanding of private keys can help prevent loss of funds and other mishaps, but it can also offer useful insights into how Bitcoin works. This guide outlines the most important private key concepts for effectively using Bitcoin.
Bitcoin: A Secure Messaging System
Although Bitcoin is best known as a payment system, underneath it all runs a secure messaging system built on the Internet. Instead of relaying emails, texts, or web pages, the Bitcoin network processes value-transfer messages called transactions. Private keys play a central role in verifying these messages, identifying senders and receivers, and in securing the network.
An example helps illustrate the problems that private keys solve. Imagine Alice wants to pay Bob 10 bitcoin (BTC). She begins by creating a transaction identifying Bob as the payee and 10 BTC as the amount to be transferred. Alice then broadcasts this transaction to all users of the Bitcoin network.
Simple Transaction. Alice’s transaction paying Bob 10 bitcoin. As it stands, Alice has no way to identify herself or Bob, nor does she have any assurance the transaction won’t be forged or tampered with.
In using this system, Alice faces two fundamental problems. First, she needs a way to identify both herself and Bob in the transaction. Alice can’t employ a central authority such as a government registry or email provider because that would create a central point of control and failure — the very thing Bitcoin was created to eliminate. Second, Alice needs a way to prevent others from changing her transaction and forging transactions in her name.
Bitcoin solves these problems through a system called public key cryptography. This system uses two pieces of information to authenticate messages. A public key identifies a sender or recipient, and can be distributed to others. A private key is used together with the public key to create an unforgeable message signature. The private key must be kept secret. Public and private keys are mathematically linked.
Alice Signs Transaction. Alice signs a transaction to Bob with her private key. Bob will later use Alice’s public key to verify Alice’s signature.
With this overview in mind, here are six things about private keys to keep in mind as you use Bitcoin.
1. A Private Key is Just a Number
A Bitcoin private key is simply an integer between one and about 1077. This may not seem like much of a selection, but for practical purposes it’s essentially infinite.
If you could process one trillion private keys per second, it would take more than one million times the age of the universe to count them all. Even worse, just enumerating these keys would consume more than the total energy output of the sun for 32 years. This vast keyspace plays a fundamental role in securing the Bitcoin network.
Because private keys contain many digits when expressed as decimal numbers, an alternative called Wallet Import Format (WIF) has been devised. This format begins with the number “5” and contains a sequence of letters and numbers. For example, here’s a private key represented in WIF format:
5KJvsngHeMpm884wtkJNzQGaCEui9HJBGFsvd3VyK5qMZXj3hS
2. Transactions are Messages Signed with a Private Key
To prevent forgery, Bitcoin requires that each transaction bear a digital signature. This signature, like a private key, is just a number selected from a very large range. Wallet software generates a signature by mathematically processing a transaction together with the correct private key.
This system works because anyone with a transaction and its signature can verify the authenticity of a message. However, a transaction signature is practically impossible to fake. The only way to produce a valid signature for a particular transaction is to use the correct private key.
Two transactions from Alice to Bob. The first transfers 10 BTC, and the second transfers 2 BTC. The same private key leads to a unique, unguessable signature for each transaction.
Unlike a physical signature you might write on a check, a transaction signature changes if the transaction changes even slightly. The way the signature will change is unpredictable, ensuring that only a person in possession of a private key can provide the correct signature.
3. Anyone Who Knows Your Private Key Can Steal Your Funds
Any transaction bearing a valid signature will be accepted by the Bitcoin network. At the same time, any person in possession of a private key can create a valid transaction. These two facts taken together mean that someone knowing only your private key can steal from you.
Many avenues are open to thieves who steal private keys. Two of the most popular are storage media and communications channels. For this reason, extreme caution must be taken whenever storing or transmitting private keys.
Software wallets usually store private keys in a “wallet file” on the main hard drive. Wallets often place this file in a standard, well-known directory, making it an ideal target bitcoin-specific malware.
To counter this threat, software wallets offer an option to encrypt the wallet file. Any attacker gaining access to your wallet file would then need to decrypt it. The difficulty of doing so depends on the quality of the encryption and strength of the password being used. Wallet files can be encrypted on many software wallets by adding a password.
Although wallet backups are a good idea, they can potentially leak private keys. For example, it may be tempting to save a backup of your software wallet to a cloud storage service such as Dropbox. However, anyone capable of viewing this backup online (which could be a surprisingly long list of people) would be in a position to steal some or all of your funds. A similar problem could arise through emailing backups to yourself or leaving a paper wallet around the house. Encryption can reduce the risk, but not eliminate it altogether.
Preventing the accidental release of private keys is the main purpose of “cold storage.” For more information, see A Gentle Introduction to Bitcoin Cold Storage.
4. Addresses are Derived from Public Keys, Which are Themselves Derived from Private Keys
A Bitcoin public key is obtained by applying a well-defined set of mathematical operations, defined through Elliptic Curve Cryptography (ECC), to a private key. Like a private key, a public key is simply a very large number.
SVG Image
Private Key to Address. A private key, which is just a number such as 42, can be transformed mathematically into a public key. A public key is then transformed into an address. Each step is irreversible.
The relationship between private keys and public keys is an example of a mathematical trapdoor - a function that’s easy to perform in one direction, but practically impossible to perform in the opposite direction. This unidirectionality lies at the center of Bitcoin’s security model.
Just as private keys can be shortened to make them more usable with displays and keyboards, so too can public keys. An address results from applying a multi-step transformation to a public key. This produces a string of text and digits, usually starting with the number “1”.
Notice that no network is needed at any point in the generation of a private key or the corresponding address. Every computer on the Bitcoin network knows about the mathematical relationship between public and private keys. This enables each participant to select private keys and sign transactions independently of the Bitcoin network. The enormous private keyspace ensures that any properly-selected key will be unique.
5. Security Depends on Choosing a Good Private Key
Knowledge of a private key is the only verification needed to spend funds from a Bitcoin address. Private keys should therefore be kept secret. However, careless selection of a private key can lead to theft just as easily as its accidental release.
For example, imagine that we want to use a private key that’s easy to remember. The number 1 is both easy to remember and a valid Bitcoin private key. But how secure would it be?
The private key 1 generates this address:
1EHNa6Q4Jz2uvNExL497mE43ikXhwF6kZm
If you follow the link, you’ll notice that the address has already been involved in over 1,000 transactions for a total of over 7 BTC within the last few years. If you wanted, you could easily spend any available funds at this address because the private key is known to you.
Now imagine you’re a thief determined to steal bitcoin. One strategy might be to compile a list of easy-to-remember private keys. Next, generate the addresses for these keys and monitor the Bitcoin network for incoming payments to one of them. When one arrives, immediately sign a transaction moving the funds to another address you control.
Contrast the ease of this scheme with a situation in which a private key was chosen by a perfect random number generator. With no clue what the key might be, brute force iteration would be the only option. As we’ve already seen, carrying out this plan is physically impossible.
What would happen if the random number generator were not quite random? For example, what if all output private keys were clustered about a constant value within a narrow range?
Private Keyspace. Random private key distribution (left) versus one that is clustered (right). The clustered distribution limits the search space, favoring an attacker.
Any attacker aware of such a defect could drastically reduce the necessary search space. Under the right conditions, it would become practical to monitor all of the addresses based on the faulty random number generator and steal funds from any one of them at will.
The need to select a good private key becomes especially important with brain wallets. One method to create a brain wallet starts with a passphrase such as “to be or not to be”, then applies a mathematical function to convert this text to a private key. Applying the most popular conversion algorithm (SHA-256) to this passphrase generates the address:
1J3m4nneGFppRjx6qv92qyz7EsMVdLfr8R
If you go to blockexplorerer and search this address, you can see that this address was used recently to store funds, which were immediately withdrawn.
Unfortunately, it’s not always easy to tell what qualifies as an insecure brain wallet passphrase and what doesn’t. Attackers can exploit this uncertainty and the inexperience of new users to steal funds. For example, a thief might compile an enormous database of common phrases and passwords. Such a database might number in the trillions of entries or more, but would still be searchable in its entirety with little computational effort.
Compare this situation to the one with website passwords. If you register for a web service using a password someone else happens to have chosen, you don’t take over their account because your username must be unique. Bitcoin private keys are different in that they serve the dual role of user identification (via address generation) and authentication (via digital signatures).
Secure private keys are generated with a high degree of unpredictability so they can’t be guessed before or after the fact.
6. Private Keys are (Somewhat) Portable
For the most part, wallet software hides the process of generating, using, and storing private keys. However, private keys can become visible from time to time. When this happens, understanding private keys and how they interact with your specific software becomes important.
Paper wallets present the most common route by which private keys show up outside of software wallets. Although they come in a multitude of formats, the essential feature of any paper wallet is a printed private key.
Support for using externally-generated private keys varies greatly across wallet applications. For example, private keys imported into Armory are not preserved in future backup recoveries using the most common and recommended procedure. Likewise, importing private keys into MultiBit changes the behavior of that wallet with respect to change addresses.
Should one wallet application begin to malfunction, its private keys can be imported into another application. This rescue procedure provides the second main route through which private keys become visible to end users. A closely-related procedure consists of restoring the state of a software wallet through a backup file.
Before losing funds due to preventable mistakes, understand how your software treats externally-created private keys — before importing them. Regardless of the specific wallet application being use, private keys kept or maintained outside of a software wallet need to be handled with care to prevent loss and theft.
Conclusions
Bitcoin can be thought of as an open messaging system secured by public key cryptography. In contrast to other systems protected by username and password logins, Bitcoin is secured through digital message signatures created with a unique private key. This single point of access places a very high value on the secure generation, use, and storage of private keys.