During the National Day holiday, Foresight News selects an in-depth article for you every day.
Written by: Joe Zhou, Foresight News
This article was published on September 23
"Of ten quantitative teams, eight lost money, and one or two even liquidated their positions and disappeared."
This is a rough impression left by Lao K after investing in more than a dozen quantitative teams in the currency circle last cycle.
Of course, this is not an industry statistic. This is just the result that a person who has been involved in the encryption industry for a long time has seen after investing in more than a dozen teams with real money.
Old K is a partner of a Hong Kong quantitative trading team that I met recently in Bali. In 2016, he started to get involved with cryptocurrency and officially entered the industry in 2017. In the past cycle, he not only invested in more than a dozen quantitative teams in the currency circle, but also in crypto VCs such as Waterdrop Capital and Hack VC.
In the past nine years, he has experienced almost all the crazy moments in the currency circle. When the Yi coin was still 1 yuan, he bought it, and later it reached a maximum price of nearly 1,400 yuan. When BNB was still 1 yuan, he and his team also bought 50,000 BNB, and now the price has reached about 800 US dollars.
But compared to these skyrocketing stories, he wanted to tell me another thing: making money may not be the most difficult thing in quantification in the currency circle, but surviving is.
In order to have a deeper understanding of quantitative trading, last cycle, he invested real money in more than a dozen quantitative teams, hoping to see how others made money.
The results surprised him: roughly speaking, eight out of ten companies were losing money, and some had even disappeared.
Some people were addicted to contracts and eventually had their positions liquidated, some had their funds stolen from the chain, some had exchanges go bankrupt, and some suffered heavy losses in the collapse of altcoins.
Old K used real money to uncover a little-known corner of quantitative trading, and also allowed us to see a fact that is often hidden by the words "quantification", "arbitrage" and "low risk": in this industry, the number of teams that can truly survive the cycle may be much less than imagined.
The following is Old K’s self-report.
The first pot of gold: from the thousand-fold myth to manual "moving bricks"
In 2016, when Lao K first came into contact with cryptocurrency, the industry was still in an extremely early and barbaric stage.
In the second half of 2016, he participated in the ICO of an early domestic public chain project - Xiaoyi Coin, which later became NEO.
The participation method at that time was very primitive: the project party published the address on the website, and the investors transferred Bitcoins and then received the project tokens. He remembered that the project raised 20 million yuan in the second round and only accepted Bitcoin. The converted issuance price was about 1 yuan.
The next year, the market went absolutely crazy.
In 2017, ICOs began to explode on a large scale, and the crypto market ushered in a bull market. Especially during the surge after the "94 Incident", the Yi coins in Old K's hands experienced an astonishing increase - from the original price of about 1 yuan to a maximum of between 1,200 and 1,400 yuan.
The market price of nearly a thousand times made him truly feel the explosive power of this market for the first time.
But what really allowed the old K team to make real money and realize that there were "business opportunities" was not the thousand-fold coin, but a simpler phenomenon: the price of the same coin on different trading platforms was actually different.
It was extremely early in the market, and information, liquidity, and prices between exchanges were not completely synchronized. The same coin sold for 100 yuan on platform A may already be sold for 105 yuan on platform B.
After the bull market started in 2017, Lao K and others already held some coins and quickly discovered the price differences between various platforms. They do their own research and manually operate the currency across platforms: buy wherever it is cheap and sell wherever it is expensive.
There is no complex quantitative system and no mature trading infrastructure. Tracking the market, calculating spreads, transferring coins, and placing orders are all done manually.
"In just one month, I made back my principal of 100,000."
This experience made Lao K realize for the first time that what he was really interested in might not be betting on how much a certain currency would rise, but whether he could continue to make money from the price difference in the market itself.
As the transaction volume increases, manual brick moving will soon become overwhelming.
They started buying other people's trading software. Later, a month's software usage fee even cost tens of thousands of yuan.
Old K did some calculations: Since it costs so much money every month, tens of thousands of dollars is enough to hire a few people to develop it by ourselves, why not write one by ourselves?
So, a team of only 7 people started to develop their own quantitative trading platform.
Originally, they wanted to open the platform to external users, but due to policy reasons, it eventually became purely internal use. And this system has been iterated to this day.
In addition to Yi coins and early brick-moving, what impressed Lao K the most was BNB.
In July 2017, Binance held an ICO, and the subscription price was also about 1 yuan, settled in ETH. At that time, the threshold for participation was very low, and you only needed an email address to register. Lao K’s team used three accounts, and each account subscribed for more than 10,000 BNB.
Later, this batch of BNB was sold in batches starting from 1 yuan and sold to 80 yuan, earning an average of twenty or thirty times. At the time, this was already a considerable profit.
But looking back from today's perspective, this transaction still leaves a little regret: BNB later rose to a maximum of more than 800 US dollars. In order to maintain the Binance VIP level, Lao K later bought some back at around $100.
He has no regrets: "You can't make all the money."
This sentence may also explain why he did not continue to look for the next "thousand-fold coin" later.
Rather than betting on the rise and fall of a certain asset, he became more and more interested in another thing: whether he could find a way to continue making money that did not rely on the market direction, but on the market structure itself.
The year was 2017.
Lao K himself may not have expected that a seemingly accidental "moving bricks" would eventually turn into a career that lasted nine years.
From earning 100% monthly to 20% annually: Quantitative money is getting harder and harder to make
In 2017, it doubled in one month.
In 2021, it doubled in one year.
By 2026, the annualized rate is 20% to 30%, which is already a "good" income in the eyes of Lao K.
Over the past nine years, in the same industry, the rate of return has followed a completely opposite curve. Why?
Old K lamented: The era when it was easiest to make money through quantification is passing.
When I first started doing quantification in 2017, the market was still very early. Information, liquidity, and prices between exchanges are not completely synchronized. As long as you find the price difference, you may make considerable profits.
Old K recalled that the team was able to double its principal in one month by doing cross-exchange arbitrage in the early days. That was not because the strategy was complex, but because the market was inefficient and there were few competitors at that time. Many opportunities that have been quickly eliminated by programs today could even be caught manually back then.
But the good times didn't last long.
The market entered a bear market in 2018, with Bitcoin falling from nearly $20,000 to more than $3,000.
For ordinary investors, the biggest fear is falling currency prices.
But for the arbitrage team, the scariest thing is not even the decline, but the lack of trading.
Without trading volume, it is difficult for spreads to form; without volatility, there is no profit for the strategy.
A bear market does not necessarily mean losing money, but there may be no money to be made at all.
In those years, the team’s previously accumulated altcoin assets shrank significantly, losing about 80% at one point. Fortunately, there were only 7 people at the time, and the cost was very low. Everyone received a basic salary and continued to invest the money they earned in server and system operations, so we got through it bit by bit.
In 2021, quantification will usher in another golden age. The crazy rise of altcoins has brought unprecedented liquidity and volatility, and almost all strategies such as arbitrage, market making, and trend trading have their own opportunities.
Old K still remembers a legendary case of market making in the industry: one day, it doubled 15 times.
There is a team that relies on a high-frequency market-making strategy to achieve such profits in extreme market conditions.
Of course, this kind of income cannot be easily copied to large funds, but it at least shows that in the market of that era, there were indeed opportunities for huge profits that are difficult to imagine today.
Later, Lao K realized more and more that what quantification really liked was not just the bull market.
Quantification likes "inefficiency".
As long as the market is chaotic, dispersed, and immature enough, price differences will occur; and when more and more funds, algorithms, and institutions enter, the market becomes more and more efficient, and the originally quantified profits will naturally become thinner and thinner.
After 2022, the market will enter a downturn again, but Old K discovered that this time it was not the bear market that really changed the industry.
It's that markets are becoming more and more efficient.
The price difference that existed in the past few minutes or even dozens of minutes may now be smoothed out by the program in a few milliseconds.
In the past, a 100,000 yuan account and a few people who knew code could find many opportunities; now, behind the same opportunities are more and more professional institutions, low-latency systems and algorithms.
It’s not that quantification can no longer make money, but that everyone has learned how to make money.
Old K’s team also began to adjust. They gradually reduced their altcoin positions and replaced some of their quantitative bottom positions with BTC and ETH, which have better liquidity and stronger market consensus.
By 2026, he feels that this market round is more like a local market for Bitcoin, and it is difficult for the liquidity and financial support of altcoins to return to the past.
After nine years, Lao K’s yield has also drawn a clear downward curve:
2017: Doubled in one month.
2021: Double in one year.
2026: annualized rate of 20% to 30%.
The numbers seem to be getting smaller, but the competition behind them is getting fiercer.
In the past, making money relied on the immaturity of the market; now, making money depends on you being faster and more stable than others, and not dying.
This does not mean that quantification fails. Quite the contrary, it means that the market has gone from an inefficient wasteland to an increasingly crowded poker table.
The market is becoming more and more efficient, huge profits are becoming less and less, and quantification is becoming more and more like a real financial business.
Nine years later, Old K is still doing the same thing: looking for price differences in the market.
It’s just that the era of doubling the monthly income has passed.
With no money to make, they started looking for the next "pond"
Quantitative teams are rarely committed to one asset.
Wherever there is liquidity, wherever there is volatility, and wherever there are price differences that others have not discovered, they will go wherever they go.
The same goes for Old K’s team.
In the past nine years, they have experienced three obvious migrations: from the mainland to Hong Kong, from arbitrage to market making, and from altcoins to tokenized US stocks.
The first time was in a different place.
After 2022, as Hong Kong re-embraces Web3, Lao K began to seriously consider a question: If quantitative trading will continue for another ten years, where should the company be placed?
In the end, they chose Hong Kong. They registered companies, settled in the Cyberport, and successively solved their identity problems through talents, professionals, etc. This year, Old K even sold the house in his hometown, completely cutting off his retreat.
For him, the biggest attraction of Hong Kong is not a specific policy, but three words: safety, compliance and certainty.
In the past, quantification was more like a few people sitting in front of a computer doing transactions; but when the team decides to operate for a long time, the company's registration location, bank account, fund custody and team recruitment all become problems that must be solved.
For the first time, they realized that quantitative trading is not just a trading strategy, but also a financial business that requires long-term operation.
The second time is to change the way of making money.
In 2017, they started by arbitraging across exchanges: buying wherever it was cheap and selling wherever it was expensive.
But as the market becomes more and more crowded, it will be difficult for this method to replicate past returns. Old K described it very directly: "The top 10% of institutions eat up 90% of the profits, and the remaining 90% of institutions can only grab less than 10% of the leftovers."
As a result, they began to expand from arbitrage to market making, trend and other strategies, and also began to try new tools such as AI.
But what really alarmed them was the altcoin market they were most familiar with.
By the first half of 2026, Old K clearly felt that the liquidity of altcoins was shrinking. There is less trading volume, less funds, and fewer opportunities to accommodate strategies.
He said a very vivid saying: "When the water in the pond dries up, the fish will naturally disappear."
So, they started looking for the next "pond."
This time, they are eyeing tokenized U.S. stocks.
In June this year, the team first spent $10,000 to test a tokenized U.S. stock strategy.
Three months later, that funding had grown to $300,000.
Although this strategy currently accounts for less than one-tenth of the team's total funds, in Old K's view, it represents a new direction. In the past, they traded crypto-native assets; now, they are beginning to study tokenized versions of traditional U.S. stocks such as Tesla and Nvidia.
The biggest attraction of this type of asset to them is that on the one hand, it is the price anchor of the traditional financial market, and on the other hand, it is the 24/7 trading feature of the crypto market.
As a result, the team began to re-develop trend indicators suitable for all-weather trading, trying to transfer the trading experience accumulated in the past nine years to a new market.
Looking back, the changes in this team are actually very interesting.
In 2017, they were looking for price differences in the altcoin market;
In 2022, they moved the company to Hong Kong to find a more certain operating environment;
In 2026, they began to look for the next "pond" with liquidity, volatility, and opportunities.
Nine years later, it’s not just the strategy that has changed.
The market has changed, the assets have changed, and the city has changed, but their logic of looking for "inefficient markets" has not changed.
This may be the most realistic way for the quant team to move through the cycle: instead of waiting for a market to come back, wait for the market to run out of water and then look for the next pond.
I have made 1,000 times my profit, but I still can’t escape FTX and FCoin
After nine years of quantification, Lao K has experienced almost all types of extreme market conditions: "94" in 2017, "312" in 2020, "519" in 2021, and "1011" in October 2025.
For ordinary investors, these days often mean plunges and losses; but for quantitative teams, extreme market conditions sometimes mean opportunities.
Others are afraid of fluctuations, but sometimes the quantitative team is waiting for fluctuations.
Because the more violent the market, the greater the price deviation between different exchanges. The price difference is only 0.1% under normal circumstances, but may quickly expand to 1% or more in extreme conditions. For arbitrage strategies, this means more profits to capture.
The bull market in 2017, the “312” in 2020, and several subsequent extreme market conditions have all brought a large number of arbitrage opportunities to Lao K’s team.
But the problem lies precisely here. Quantifying the money earned is based on the normal operation of the entire trading system.
When something went wrong with FCoin in 2018, Lao K’s team directly lost about 20% of its principal.
This was not a strategic loss. If a strategy loses 20%, you can at least know why you lost; but if there is a problem on the exchange, you may not even have made any mistakes and your money will be gone.
A few years later, something similar happened again.
When FTX crashed in 2022, the old K team had already increased their risk awareness and did not put much money on FTX, ultimately losing only $80,000.
But FTX had an even greater impact on him.
Because FCoin was just a rapidly rising exchange at the time, and FTX was already one of the leading platforms in the industry. Many professional institutions have believed that placing funds on such a platform is a relatively safe choice.
But in the end, it turns out that scale, brand and professional image cannot be equated to financial security.
This also changed the way Old K later judged the exchange. In the past, he first looked at handling fees, depth and API stability; now, he also asks: Who holds the money? Who regulates the platform? Are customer assets and platform assets truly isolated?
Moreover, quantitative teams like Lao K usually do not concentrate their funds on one exchange, but disperse them across multiple platforms. Binance, OKX, Bitget, Gate, Coinbase, KuCoin, Hyperliquid, Lao K said that these platforms now have funds, but the allocation ratios of different platforms are different.
Because he is increasingly aware that: you think you are doing arbitrage, but in fact you may be taking credit risks for the exchange.
This is also the most contradictory part of quantitative trading.
In theory, arbitrage strategies can minimize directional risk - you don't need to judge whether BTC will rise or fall tomorrow. But in order to execute the strategy, you must put funds into the exchange.
Therefore, what quantification really needs to manage is never just the model.
The model solves "how to make money", while the risk control solves "whether the money earned can be kept."
When he first started doing arbitrage nine years ago, what he was most concerned about was where there was a price difference. Nine years later, he's still looking for the difference.
But now, before putting money in, he will think about one thing first: If this exchange disappears tomorrow, what will happen to my money?
This may be the biggest change after nine years of quantification.
Ten teams, eight losses: the "quantitative paradox" hidden in the siege
Old K conducted an “experiment” with his own funds.
In the past few years, he has provided funds to more than a dozen external quant teams through API sub-accounts, ranging from US$10,000 to US$100,000 per transaction.
His original intention was simple: on the one hand, he wanted to communicate with his peers to see if there were any new strategies worth learning from; on the other hand, he also wanted to make some financial gains along the way.
The two parties agreed that if money was made, Lao K would give the team 20% to 30% of the profit share; if there was a loss, Lao K would bear all the losses himself.
But after a round of investing, the result was beyond his expectation: among more than a dozen teams, about 80% lost money in the end.
Some retrace 20% to 30% and then stop the loss urgently, and some even liquidate their positions directly to zero. There are only a handful of companies that can truly produce positive returns stably in the long term.
This experiment with real money also allowed Old K to see clearly a paradox hidden in the quantitative industry: truly top-notch quantitative teams that can survive bulls and bears are often not short of money.
They have their own funds and stable strategic capacity, and do not have strong external financing needs at all. More importantly, even if you have money, you may not be able to invest it.
On the contrary, there are teams that are easy to access and willing to accept external funds. Many of them are small in scale, and their strategies and risk control are not mature enough.
The easier it is to invest, the less likely it is that you really want to invest; the one that is really worth investing may not be your turn at all.
How on earth do these teams lose money?
Old K discovered that many people mistakenly believe that "quantification" means guaranteed profits without losing money. But in fact, quantification can only program trading rules, but it cannot eliminate risks.
In order to pursue high returns, some teams add leverage to altcoins, which may result in instant liquidation when encountering extreme market conditions; other teams do stablecoin revolving loans and continuously add leverage. Once they encounter a black swan like "1011", their profits and even principal in the past one or two years may be reduced to zero all at once.
Therefore, when Old K later judged a quantitative team, he mainly looked at four things:
First, running time. Have you ever experienced extreme market conditions such as 312, 519, and 1011?
Second, yield. A 100% annualized strategy may sound tempting, but it often has capacity constraints and is difficult to sustain over the long term.
Third, the maximum retracement. At the same rate of return, the smaller the retracement, the healthier the capital curve is usually.
Fourth, management fees. In the end, management fees will be deducted from the actual income.
"Of course, when we say 'Ten quantifications and eight losses', we refer to ordinary people or the quantitative teams we come into contact with on a daily basis." Lao K emphasized.
Over the past nine years, he has seen coins multiply thousands of times many times, and he has seen exchanges collapse; he has seen market-making funds multiply 15 times in one day, and he has seen teams liquidate their positions and return to zero.
Quantification is neither a myth nor a scam. It just makes the way of making and losing money more procedural.
It is not difficult to make quick money by luck. What is difficult is to survive in this market for nine years.