Professor Boring and his Fantabulous Scan-traption–and why I’m not a Systems Trader

February 28th 2011 at the UC Irvine Student Center — The Coastal Trade Securities Q&A

We were about to have our greatest event yet. I had invited three prop traders from Coastal Trade Securities to do a Q&A presentation for our trading club. Two of them, Ed and David, were ex-NYC traders turned firm managers. They hyped up the third guy, John, as their “chief technical analyst”–basically the lead dog who sets the game plan for morning. Fun fact, John also moonlighted as a minor league hockey ref.¹I have my theory that it takes a very peculiar person who cares a LOT about the rules to become a referee–which is an unglamorous and often hated figure in sports. When you learn more about John… maybe it’ll make more sense.

Ed and David kicked things off with a quick spiel about being a prop trader at Coastal Trade, aiming to hook the students in the audience who might dream of a trading career. Then, they handed the mic to John for a PowerPoint presentation detailing his highly intricate moving average trading system.

Big mistake.

About 30 minutes into John’s talk, the room started hemorrhaging attendees. One by one, students got up and quietly shuffled out the door. It was brutal. As the club president who painstakingly organized this “landmark” event—the first visit from a legit prop trading firm to our campus—I tried to bribe the early leavers to stay with promises of complimentary pizza on the way. Imagine John as your least favorite math teacher dryly dissecting a lesson until it dies–listing rote example after rote example with no inflection and no idea when to stop. There were also moments he would try to break down a trade and it felt like he was sneering at us–look at how much I know and you don’t!

Nobody gave a shit. The Q&A didn’t last very long.

That’s why John is Professor Boring. He bored the piss out of the crowd at my biggest event.

Drudgery aside, I could tell John poured his soul into knowing the in’s and out’s of his trading system. I’ll give him that. Ed and David offered me a spot to trade with Coastal but I declined since I had already committed to MBC Securities. Then I graduated, moved, traded in NYC for years, and I never thought about him again…

Reuniting on Twitter

…until 2016. Although I didn’t know it.

The twitter ecosystem of finance experts²(or FinTwit, as some call it) had grown exponentially since 2011. Every trading guru wanted to make their impression with gaudy PnL or lifestyle porn like sports cars and luxury watches. Almost all of them repulsed me except one… I observed this one particular account, twitter handle @HowStocksMove, who would consistently post impressive and realistic trading execution screenshots. He kept it simple and often kept to 1-2 entries & exits at most at a time when the “en-vogue” daytrading thing to do was to scalp within a position as often as possible. His execution points made intuitive sense to me. The tweets accompanying his charts were often infuriatingly pithy, like this:

HowStocksMove’s entry and exit points resembled what I would do–waiting for ranges to break or waiting for trend confirmation. ‘Game recognize game’, so they say. One day, I decide to DM the guy… how do you do it?

We have some productive conversations and just like that, I decide to make the leap–join his chatroom and learn from some of his videos³for a small fee of course. His voice sounds eerily familiar. I would then learn certain facts about him throughout my first few months in chat–that he was a FINRA registered prop trader⁴fun fact: Coastal was acquired by Y5 Capital in 2013 so PB was technically registered as a Y5 trader, just like I once was who lived in Orange County–and then I put two and two together. Then when I felt 100% sure, I asked him–you’re that guy aren’t you!? From the Coastal Trade Q&A at UCI? And indeed, he confirmed as much.⁵we had never actually spoken to each other then, I had only spoken to Ed and David. PB had no idea I was the club founder/president who planned the entire event @HowStocksMove turned out to be Professor Boring himself–the same guy I met five years ago who set the whole damn room to sleep. It’s one of those movie-like moments that feels too on-the-nose to be believable.

Surprisingly, the 2016 version of the Professor wasn’t nearly as monotonous as the guy who presented at UCI five years ago. Now he had a bit of a smart-assed edge to how he carried himself online. He also had this eclectic set of metaphorical phrases to repeat to his students–pearls of trading wisdom that later I will refer to as ‘PB-isms’. He embodied the sarcastic wise old teacher trope, trying to instill upon us discipline and patience.⁶even though he wasn’t that old. he was around 5 years older than me and had been trading since 2005. So late 30’s around 2016.

How Stocks Move 1.0

The Professor broke down trading into two simple concepts–Context and Timing. Context describes the bigger picture concept you are trying to capture edge from and the Professor had four distinct contexts: earnings breakout⁷(trading in the direction of the earnings gap), earnings reversal⁸(trading against the direction of the earnings gap), overextended PoS stock⁹(a stock, usually a micro-cap, going parabolic 1000% and its many variations. PoS means piece of shit, obviously), and fluctuation¹⁰(good chart setup but otherwise nothing special, this was the weakest context). Timing refers to the entry pattern, which he breaks down into granular units of price data.

The first concept to understand is the bar break (bb). This occurs when a bar breaks the high or low of the previous bar. Bar breaks exist on any timeframe on you want.

The second concept to understand is bar break against bias (BaB). If your bias is long, a bar break against your bias would be a bar break to the downside. If your bias is short, a bar break against your bias would be a bar break to the upside.

This is needed to understand the 3rd concept that combines both the 1st and 2nd concept together to create price formations that PB called cups ¹¹i.e. cup and handle pattern in classic technical analysis. A cup is basically a pre-determined sequence of BBs and BaBs, in conjunction with a moving average, that creates a clearly defined price trigger with a clearly defined stop. Once this price gets above trigger, the cup is now ‘valid’ for both entry and stop¹²(usually the high or low of said cup’s range). This is the Professor’s primary timing technique.¹³(imagine buying only green bars without any BaB’s. there are no ranges and no clearly defined support/resistance levels. you are in effect, just chasing pure strength.)

I had never before encountered any guru who could define his confirmation signals and entries/exits to an exact price. Too often, you get a guru who’s a bit more all over the place. Scaling into washout, watching $10 level of interest, scaling in more, giving it room, want to see if it fails as mental stop–that’s the kind of messy commentary that will overwhelm or confuse any new trader.¹⁴I’m not saying this can’t work. I know plenty of great traders who don’t define everything to a penny. I’m just saying this is harder to learn from because it’s messy and unclear. PB would call this ‘not knowing exactly when to push the button’

On his favorite contexts, PB would hold for a pre-determined target or until the closing bell. On fluctuation contexts, he’d sell into strength or cover into weakness to lock in profits quickly. He posted his executions at the end of day to show that he was a man who wasn’t all talk and that he practiced what he preached. He posted both wins and losses and maintained intellectual integrity at each step.

This is the HSM method: You pick your context. You wait patiently for your setup–usually a cup or a higher timeframe bar break. You enter and set a stop, wait for target. It was very simple and structured. You could see the appeal. Well-defined trigger/stops = less guess work = reduce over trading.

The professor’s academic style of teaching and lack of any kind of hypey marketing attracted hard-working, humble people who wanted to learn how to trade, as opposed to any unserious dilettantes who wanted to get rich quickly. HSM traders all traded their own way–trends, mean reversion, large cap, small cap–but we were all on the same page with technicals, via using price and cups as our primary timing mechanism. It was a nice place to be compared to the usual chatroom with 100 different traders spazzing out 100 random callouts.

A Different Flavor of Guru

Like any good trading guru, you have to have your own unique terminology and catchphrases–what I call the PB-isms.

TGPP which means “Trade Good, Patience Pays” — he always signed off his watchlists and daily recaps with TGPP. The Professor focused on process first. Get that right, the money will come.

“Learn to read what the author is writing.” — the so-called author is a reference to the greater forces that drive price action on a chart. It’s one of the PB-isms he repeats the most. He’d constantly refer to “the author” letting make us money or at times, not letting us money, and it was just something we had to accept and could not control.

Is it a Hippopotamus? — this question is like a mental test he made up for his students to instill trigger discipline. Is there full setup and a trigger price? If yes, then trade it. If no, then wait. But instead of saying that, he would ask is this a Hippo? Along side, he then posts two images–a hippo and an animal that is clearly not a hippo. You probably don’t get it and I don’t blame you–one of those things where you had to be there.

WWHSMD? –stands for “What would HSM do?” — there were days where the Professor just didn’t want to do anything–answer student questions, trade recaps, post videos–so he came up with this phrase to guide students while on break.

🐢 and any turtle references — at one point, a bunch of traders in chat adapted a turtle emoji on their social media handle. The Professor saw himself as a Richard Dennis or William Eckhardt-like mentor, sharing his system with his students to ensure their success, much like the 1980s Turtle Trading Experiment.¹⁵PB almost certainly has read the Market Wizards series cover-to-cover multiple times–hold this fact in your mind because I will reference a passage from William Eckhardt himself later.

The Professor could often be the embodiment of the 🤓emoji, particularly when giving his opinion on market factors that lay outside of his system. He thinks he’s above considering things like fundamentals or market cycle in his trading. The best way to wind him up would be to ask him if “tape reading” was necessary for market timing.¹⁶he thinks it’s worthless One of his most controversial tweets was that nobody ever really needs to truly “adapt” in the market¹⁷I would later learn that this stance would come off as a little pedantic, as he would be making endless changes to his trading system, but he instead calls these changes ‘refinements’.

I think of the Professor as the Anti-Guru of Guru’s. Sure, he taught trading and charged for it, but he avoided all the obnoxious guru antics of the time, which was a breath of fresh air. He didn’t indulge in lifestyle porn on Instagram. He didn’t constantly post about a holiday sale for his chatroom price. He didn’t brag about PnL. Aside from a snarky know-it-all tweet once a week, he kept it low key.

For students who weren’t yet CPTs, PB was the only true authentic guru and they wanted to emulate him almost 100%. Personally, I viewed him more like a trading coach. I didn’t agree with his every opinion or trade exactly like him, but that’s okay. He added value and he cared about making traders better.

You want to know what my biggest takeaway was from the HSM 1.0 era? It was to stop wasting energy on irrelevant bullshit and to focus on a good structure for consistently buying low and selling higher. That’s all trading is. I stopped caring about things like edge erosion because trends and mean reversion are timeless concepts. I stopped caring about a stock being ‘too crowded’¹⁸it is until it isn’t. I stopped trying to do deep research or fancy volume analysis. I stopped caring about ‘algo manipulation’ and started to wonder if it was just a made up term to excuse bad trading.

I did have some personal breakthroughs that I credited to the trading patterns I learned from the Professor. I used to struggle a lot with broad market volatility plays. In the February 2018 VIX-blowout, I saw one of PB’s cup patterns pre-market on VXX. It triggered, I got short, and I held for a big opening move.

During these days, I would DM’d the Professor all the time. I was profitable but I had overtrading issues to resolve. My training at MBC taught me to stare at the screen all day and trade as often as possible. I wanted to de-escalate that desire to hyper trade everything and simplify my process. Something he said stuck with me.

The Professor: Stocks move how they move. You can make it as complicated as you wish.

I had no idea what was about to come.

How Stocks Move 2.0 — The Introduction of the Fantabulous Scan-traption

Sometime later in 2018, the Professor made this miraculous discovery called “data mining.” His goal was to get closer to systematic trading where every decision follows a fixed rule based on backtested data. The most important question of the trading day is such: What are you going to trade? And for the Professor, he wanted that question answered objectively–with numbers instead of opinions.¹⁹the old HSM 1.0 process on ‘what to trade today’ was to create a watchlist based on looking at briefing news, looking at charts, and then labeling each stock under a specific context

Thus the Fantabulous Scan-traption was born.

To build his scan, he began with a clear goal: how can I quantify the setups that consistently lead to the most impressive “chart porn” outcomes?²⁰chart porn definition: obvious daily chart context, followed by trigger, followed by clean move to target He started by identifying trades that he believed met a specific aesthetic ideal—those with standout outcomes from a trigger, like fading all day to a distant target–and then label them under a specific trading context²¹for example: gapper, PWR, multi-day runner, or intraday spike. more on that below. For each instance, he recorded the stock symbol and date in a spreadsheet.

Next, he used Trade Ideas’ backtesting feature to perform what he called a “Data Pull.” This process generated over 25-50 data points for each stock–parameters from that specific date. Based on this dataset, he would adjust his parameters to tighten up the results.

For example, take RSI, a popular indicator of whether a stock is overbought or oversold. If the lowest RSI in his dataset was 70, he would set a minimum RSI filter of 70 for his scan. Similarly, if all the stocks in the dataset traded at least 10 million shares in prior-day volume, he would set a minimum volume filter of 10 million. Stocks that fell below these thresholds—such as 9.99 million in volume or an RSI below 70—welp too bad, they’re cut off from scan.

This process was designed to filter out mediocre setups, leaving only stocks that met his predefined criteria prior to the 9:30am open²²(scan candidates had to remain on the box before 9:30 otherwise it was not considered valid. pre-market actions would have a significant impact on filters). He repeated this refinement for each filter to create a real-time scan that would populate only with stocks matching his ideal parameters.

The next step is to manually backtest each chart from the scan’s historical data under a common entry system and measure its win rate and risk-reward. The tested entry methods he used were built on the same principles of HSM 1.0, like bar breaks and cups. One common entry method he kept using was 30min low²³(at MBC Securities, they called this an opening range breakout or ‘ORB’. there is more implied informational value placed on the range of the first 30 minutes). It’s basically a bar break low following the first 30 minute opening bar at 9:30. Example: high and low from 9:30-10am is $19-$20? Then 18.99 is the trigger price to short.²⁴(it could also be 18.89. he often used a discretionary “wiggle room” to add to his stop entry and stop exit signals, but we’ll get to that later) On other scans, he might look for a cup pattern. He also had conditional entry methods–for example, the entry would be a 30 minute low with the condition that the second 20min bar of the day was a bab against the first²⁵(reminder: this means break against bias, meaning the bar went against his biased direction and THEN it went lower for the signal. so he wants a specific sequence).

After the entry, he would also backtest which target would be hit the most frequently on average–usually they would be simple moving averages on different timeframes. If the results were sufficient, scan goes live for real trading.

Let’s try to break it down, each step.

Professor Boring wants to create short biased gapper scan, so he collects 10 charts of high % gappers that closed significantly lower than their open. Then he pulls data from all of them. It might look like this.

Then based on this data, he customizes the min/max settings on the filters to include this “ideal” data set but cut off any new candidates that cannot meet the data set’s min/max range.

Hopefully the other stocks that populate will also be winners too. These are the “factory default” settings. On a year’s worth of data, maybe this scan came up with 36 different candidates.

Now comes the backtesting. Have a basic entry method, a basic target²⁶(he used multi-timeframe moving averages) and test it. So PB manually combs through every single chart via historical charting software²⁷I personally used ThinkOrSwim’s on-demand feature, PB used eSignal to find every candidate that made a 30 minute low. Of those candidates, he logs the winners and the losers based on his stop/target criteria.

How does a scan get the rubber stamp for live trading? Good results. Say PB observes that there were 30/50 candidates that triggered and of the 25 triggered, 20 of them went to a winning target (or at a minimum, closed lower than the trigger price) while the other 5 were stop outs. That’s an 80% batting average. Then he’ll also measure the magnitude of the wins and compare it to the risk (which is trigger price minus the highs of the day), average them all out, and find out that winners amount to 1R on average. Great data = great scan = go live and make money.

That’s the Professor’s new vision that he called “black and white” trading. Clearly-defined contexts and trading signals based on proven historical data. Each day follows a simple script: wait for the box to populate a symbol, follow scan’s trigger/stop/target rules, make lots of money.

At first, we all LOVED the Professor’s revolutionary new take on HSM trading. We now had in our hands this data-driven method with a well-defined process and a known high win rate. This felt like the very beginning of something big.

Now everyone in chat wanted to follow PB’s footsteps and play scientist in the stock market. We were all pulling data and backtesting everything in sight and trying to come up with as many scan opportunities as possible. For whatever reason–the HSM 2.0 method had a predisposition towards developing short-biased mean reversion strategies. It was difficult to build scans with a statistical edge for different categories like event-based trading, trend/momentum trading, or long-biased mean reversion. The vast majority of PB’s scans were under the short-biased mean reversion umbrella.

Gapper (or Big Gap)— a stock up a large % in the pre-market, or in PB-ism description, “gapping out of nowhere”

Multi-day runner — a stock up multiple days in a row for an unusually large % amount

Prior-day wide range — a stock that closed up a huge % the prior day

Intraday spike — a stock that trades a huge % on an intraday basis

The general idea within these contexts were the same–that once certain stocks trade into an extreme territory, there exists a high probability that the next move would be a large correction to “the mean”. “The higher, the better”, the Professor would repeat that ad-nauseam.

Now, I understand the burning questions you might have upon discovering this secret cheat code trading strategy. Key question being… did we make millions of dollars?!

No, spoiler alert, we did not. At least not yet.

There were growing pains. There was a hint of an edge, but it was as raw as a high schooler’s science project. Common complaints included:

• Sometimes there would be a decent move but stop short of target. Then the stock would bounce and either go sideways or reverse completely. This felt devastating. Backtesting targets was not as simple as it initially appeared to be.
• Sometimes trigger felt a little late to capture what one would eyeball as the turning point of the stock. I started calling his 30min low signals cookie cutter entries because they were lagging signals that were only useful for backtesting purposes. PB called these ‘wide triggers’, referring to the wide range difference between entry and stop. Wide entries led to wide stops which drastically reduced potential risk-reward. There were times PB even acknowledged he wouldn’t take a trigger because it occurred too close to the target and took away all the ‘meat on the bone’.
• There were too many ‘successful’ turning points that were not easily defined within the HSM paradigm of cups and bar breaks. These outcomes were much harder to backtest because they required defining far more complex price sequences than just “bar break at high/low”. So we’d often miss out on great outcomes that didn’t play out from scan’s entry rules.
• Too many times we’d bury our heads in the sand and “trust the data”, thinking OMG, high win rate! We’d take signals to short against strength, only to get crushed and stop out above the prior highs. This is an example of adverse selection. The more “favorable opportunities” a stock wanted to give you to enter in size, the more likely it was to be a loser. The cleaner a stock would be from trigger, the more likely it would run away to target and thus not offer favorable prices to scale in. It was very easy to be too large on losers and too small on winners.
• Scans were too biased towards shorting micro-caps. This led to hidden costs such as hard-to-borrow stock shorting fees and higher slippage due to the thinner liquidity conditions. Nobody wanted to include this in their data.
• Some scans required holding overnight for expected targets. Even more hidden costs & risks–namely: A) overnight interest charges and B) the possibility of advese news events during the much thinner afterhours/pre-market period of trading²⁸to be fair, these could also go in your favor, such as a micro-cap stock offering at a steep discount to market prices
• Sometimes there would be an internal conflict between discretionary trading habits and systematic trading rules. This one hit me the hardest. More below.

I remember one stock, BLNK, hitting multiple scans. It sucked everyone in on 30min low trigger only to squeeze into the close.

I took it short as well but I stopped out well before highs while many in chat followed scan rules and took the maximum stop. This encapsulates one of the biggest problems I have with committing towards systematic trading. A skilled discretionary trader can sense internal strength against his bias well before the strength is ‘obvious’, which is when the squeeze is already under way. But under HSM 2.0 trading, your black box trading plan is to stop at the highs. That leads to a lot of cognitive dissonance. Do I follow the scan rules or follow my feelings? It’s a tough balance where, if you don’t commit to what you really want to do, you can fall into a results-oriented capture–Shoulda stopped early on the loser. Shoulda gave winner full stop. You’ll drive yourself un-poco-loco. I know I did and it was one of the moments where I knew I would never 100% commit to PB’s vision of “black and white trading”. I wanted some gray.

I remember one of the most loyal and devote HSM followers–Ray, a dock worker from New Jersey who dreamt of being a CPT²⁹(consistently profitable trader) in order to quit hard labor forever. For years, he felt he couldn’t get over the trading hump until he met the Professor, who gave him belief again. Very earnest, sincere guy who would spend entire weekends working on new scans. Unfortunately, Ray ran into a bad spell with scan and the BLNK loss completely broke him to the point where he decided to abandon the HSM 2.0 method entirely. He decided to be a large cap scalper instead. At the time, I thought this was short-sighted… you can’t let one brief spell of bad luck demoralize you from utilizing a proven winning strategy… that’s what the poker books and the trading books teach us, right?

…right?

The end of 2019 would conclude the initial era of the Professor’s scan–interesting concept to infuse a flavor of science into the dark art of shorting parabolic stocks but ultimately a mixed bag with glaring performance issues.

HowStocks Move 3.0 — Chasing Perfection

In early 2020, the Professor started touting a new optimization method that would transform scan forever–he would systematically purge away the losses on the data set by finding what he called “outlier data”. This would finally turn scan into a well-oiled machine that would be ready to crush the markets.

PB’s new goal: find all the losers on scan’s data set and see if there are any commonly shared data points among bad trading candidates. See if this data can herd towards one end of the filter’s range–for example, say all the losers are on the extreme left end of a 60min RSI reading. We then tighten up the filters to remove this ‘bad’ outlier data–for example, if all the herded bad data is below 61.6 60min RSI, we then tighten up the minimum filter to 61.7. Boom–a large bulk of the losers are now gone from scan’s historical data (implication is this should improve future results). I’m going to break down every step again, don’t feel bad if it’s hard to follow along. It starts to get a little wonky.

First, you take your factory settings scan and pull up the historical data on spreadsheet. Then you manually comb through each chart of every single symbol/date and label each outcome under a color–red is loser, green is a marginal outcome, and blue is an ideal “stock porn” win. Then you sort every single indicator (some scans have more 50 data points, so this could take awhile!) until you can spot a pattern of red losing trades clustering around either end of a filter’s range. As long as that cluster of bad trades doesn’t also include any of ideal blue-labeled trades, you then change the filter minimum for that particular indicator to surgically remove the losing trades from the data set. Factory settings scan now transforms into hyper-optimized scan.

It’s like working backwards. He wants a certain data set–all winners and no losers–and then he works the filters until he can get as close as possible. Later on, he would even start checking real-time scan candidates in the morning to see if any of its data was on an extreme end of a “relevant filter” and then red flag those candidates.

The HSM chat, or at least the newer traders, had been scarred by trades like BLNK and multiple bad spells but now supposedly, the shitty trades have been optimized away. The Professor starting writing entire essays in his weekend update about how he has gained a “shocking new understanding” of how stocks move from his countless hours of manual data work. He had the entire chatroom eating out of his hand with excitement.

The Professor: it took all weekend and I had to cancel anniversary plans with my wife (she’ll get over it), but I have finally eliminated the losing trades from PWR LC 2B scan

I have a question for you, if you’ve somehow been reading this far–did you read the indicator title in the picture I posted above? Do you know what “distance from pivot R2(%)” is? Have you ever, in your entire life, used this trading indicator? I have no idea what it is. You may, like me, start to ask slightly more advanced questions like “wait, does this mean this indicator I’ve never heard of us is causing trades to fail? Or is it just correlation?” I even asked the Professor… there are so many data filters and you’re not even analyzing what any of them measure, how can you be so confident this is a robust method?

According to the Professor, this does not matter. All that matters is “bad data gone!”. It does not matter how you get there. Nobody challenged his claims—everyone in the chat was just a typical day trader with very little quant training, if any³⁰otherwise we’d be at Jane Street or SIG, not on some dude’s discord trading shorting micro cap garbage. As a dum-dum day trader myself, I didn’t feel qualified to publicly second guess either but I always felt something was off–that there was no way an actual quant would sign off on this work.³¹I would honestly love to hear from a real quant on this–leave comments below.

There was even an instance where the Professor had tightened up a scan filter because the outlier trades had too many extreme readings³²meaning the stock was even more overbought/parabolic than average. The Professor had, for years, preached that the best mean reversion setups were the most extreme and most ‘obvious’ parabolic setups. I could understand wanting an RSI floor and increasing the minimum filter. I absolutely could not understand creating a ceiling for the RSI on the maximum side. What happened to “the Higher, the better”? His beliefs and his scan’s refinements seemed to be at a contradiction. PB did not seem to care, “fixing” the data set was all that mattered.

This idea of trying to perfect every little aspect of scan had completely consumed the Professor. It wasn’t just manipulating the data set. He started to sub-divide his own scans and create different rule sets for every variation. You would have “Gapper Scan A” and “Gapper Scan B” that are almost identical with 49/50 shared filters but “for reasons unknown”, they had two entirely different rule sets. If one were to look ‘under the hood’ at each chart for each data set, would you find any obvious, distinguishable differences that required all this sub-categorization? No.

It was not uncommon for one ticker to appear on multiple scans and for The Professor to draft up pages of notes on which rule set you had to prioritize first. Then low and behold, the prioritized rule set ended up being terrible for that specific trade while the other rule sets worked better.

The Professor: going to have spend the weekend to see why symbol XYZ was on both BG PIP scan and BG Main when it very clearly should have only been on BG Main ³³the one where it didn’t lose based on the complicated rule set

What began as simplicity had devolved into madness.

Updates kept coming in every week and we’d get more and more rules. These rules would often be created in reaction to a tiny sample size that troubled him, like 5-10 trades that couldn’t fit the paradigm of a prior rule set. He once introduced an entirely new timeframe with new moving averages that he had never referenced before in the HSM 1.0 era–specifically the 20 minute bar and the 20min bar/20 period SMA–that would now be a factor in the trigger/target rules. But only for certain scans. Some scans needed to “respect” the 20 min trend but the other scans did not. Which ones–I don’t know anymore. It got too confusing and even worse, the Professor’s updates slowly started to become a series of cryptic riddles.

The Professor: I went through all trades on Large Cap BD1+G %-gain scan, version 3.2a, and I found that there’s a better entry than simple 30 minute low that improves win rate. Here’s a hint, it involves looking at the first two 11 minutes bars from the open as well the 69(13) SMA and seeing the failed ___ that occurs. Once that ____ occurs, only then is ____ minute low trigger “valid”.

Even though we’ve been paying for his chat and mentorship for years, he now no longer felt the need to properly update us on the hard rules of scan. We now had to do our own heavy lifting based on his esoteric hints. Even if you do manage to crack the code for the entry method, it might just change next week after a loser appears on scan and the Professor does another weekend round of optimization.

There were traders, god bless them, that actually did the work and asked for additional help in DMs, only to get stone-walled with “can’t discuss that.” As a teacher, PB had become more aloof and unapproachable than ever.³⁴can’t count how many times PB would hang over our heads that he would close chat soon–“guys I’m tired of teaching I just want to do data work and trade!” only for ‘turtles’ to worship at his feet and beg him not to do it… so then he’ll keep chat open, raise chat prices, and respond to less DMs and post less videos than ever before

Not all of us were hypnotized into believing this was the best way forward. The Professor’s weekly updates became a recurring private joke among some of the more independent-minded chat regulars. If a non-scan candidate unwinded perfectly and made PB’s scans look totally useless? Somehow, someway… you knew what his update would be at 4pm.

The Professor: after going through the data, ticker X will be now a scan candidate going forward

Awesome, now I can make yesterday’s money.

If a scan trade fails… you just knew.

The Professor: after going through the data, ticker Y had outliers and will not be a scan candidate going forward

Too bad eliminating a ticker from an excel spreadsheet doesn’t refund us the money we lost shorting it.

I remember two tickers in particular (and there were MANY) that caused me to lose faith in HSM 3.0. They were not losing trades. They were clear and obvious “hippos” that the Professor’s overfitted scan had missed.

There was a small biotech stock, OCGN, that ran hard into a covid emergency use announcement and the stock set up for a sell the news play. You can see this market phenomenon all the time. PB chooses not to acknowledge this–he wants trading to be strictly scan and charts. According to him, events are superfluous to the so-called “author” of the markets. I didn’t give a shit and I shorted OCGN to catch a beautiful stock porn move–classic all day fader. After the close, one of the newer traders had asked PB about why OCGN wasn’t on scan and he got extremely defensive, denying that the stock had any context to it simply because it wasn’t on his meticulously backtested scan. His attitude that the entire truth of the market lied solely within his scan–that rubbed me the wrong way. ³⁵there were other times, from his commentary on chat, where it seemed PB drank way too much of his own kool-laid and honestly entertained the idea that his scans ran the market, not the other way around. he called these his ‘theories’ that he acknowledged as unproven, but you could sense he wanted to be seen as all-knowing master of the market. I wish I saved these comments from when I was still there, but alas I can only footnote this

Then there was this marijuana junk stock called SNDL. It had a massive float that ran from 10 cents to 5 bucks over a few days. It was the shittiest company ever as well as a clear and obvious parabolic chart. It was 100% the type of stock that *should* be on the PB’s multi-day runner scan. But it wasn’t (or it was, but with weird outlier red flags or it didn’t meet some overly complicated entry criteria). Perhaps you can attribute this miss to over-optimization. I didn’t care. I hit the bids early and made six-figures. After the closing bell, the red-faced Professor wrote a goddamn 1000 word dissertation about how SNDL will never be a missed trade again, going forward. Cool beans, bro.

It hit me at that moment that if a stock wasn’t on his scan, the Professor could no longer remember how to push the button. The 2016-2018 Professor who would top-tick this shitter and then cover the lows at the close–that guy’s gone. This now-insane version of him appeared completely incapable of recognizing a hippo being a hippo without a scan.

I started to think about the best trades I had ever made… the FNMA tape-reading tapes… the bitcoin swing trades. Did I need a scan to tell me to take the trade?³⁶I also thought about how deep PB had gotten into short-biased mean reversion trading. That wasn’t his thing early on but his data work seemed to corner him even further into that niche. He wouldn’t trade earnings breakouts anymore. He didn’t trade the market. He didn’t follow trend plays. This wasn’t the direction I wanted for myself. Would data even have been possible to guide those trades? Why limit myself? History doesn’t work like that, where the best opportunities repeat themselves exactly the same way. I understand wanting data behind every decision but enough was enough. The Professor, whom I once viewed as my trading coach and an ideal role model, obsessed over being this brilliant quant and it just wasn’t doing anything for me anymore. I dialed down into what I wanted to do and then I traded my ass off, scan or no scan.

BUT DID WE MAKE MILLIONS OF DOLLARS??!

That’s the bottom line isn’t it? If you make the money, all can be forgiven. Lord knows this is the bane of my existence–that I have seen some traders with nasty habits still make crazy bank because they’re just that good at the right time. Sometimes you can have a shoddy process but still be “approximately right” in the sense that the edge has been captured, even if it was via imperfect process. Was this the case?

Well… it’s a complicated answer. Nobody knows.

Professor Boring never flashed his PnL. Not even once. I have no idea how his scan performed in terms of dollars made. Data this, data that. What about the money? I don’t think PnL is the end-all-be-all for a trading mentor like some do, but there were some frustrating periods in the HSM chat where a little positive morale could’ve helped. PB’s PnL could have been a chance to deliver a boss statement, one that says “this works, here’s proof, hang the fuck in there!”. Instead, all we got were more complicated and cryptic rule updates, more empty promises that this new data set would unlock some future period of amazing performance, and more long, quiet periods of “WWHSMD?”

It’s hard to analyze the performance on the HSM 3.0 era based on imperfect information. I’m not omniscient–I can’t know the results of every single trader on the HSM chat nor the Professor’s PnL. But I was well-connected in chat, having established friendships with many frustrated traders. I didn’t love the vibes.

PB liked to compare his HSM school to the 1980’s Turtles Trading school. That was a generational trading story that produced Jerry Parker, a trading billionaire at Chesapeake Capital and Liz Cheval, who managed $150 million at EMC Capital, among many other public success stories. Those results can speak for themselves. Was there ONE notable trader, who adhered to the HSM 3.0 school of trading, that I had discovered, that either 1) made a clear leap to self-sustaining CPT or 7-8 figure trading monster?

No. I didn’t find one. Some small sub-PDT accounts did well on a %-basis–which, fine, I concede that scan trading is better than random for beginners. Some traders did alright trading a hybrid style, using a scan only as a tool and executing on their own discretionary trading–not sure if that counts.³⁷I always felt that “trading good” was most essential while the Professor stopped teaching how to actually trade. The biggest traders I know are guys who know how to freakin trade everything. Tinkering with the data set will hit diminishing returns unless you actually know how to automate the strategies. Some traders hit the fringe of being a CPT but you didn’t know if they could sustain it past the covid years. I just didn’t hear about any exceptional stories that would blow me away to the point where I had to re-evaluate my stance. From my memory, there were far too many dry spells and too many inconsistent periods. Scan would miss great trades and get into too many mediocre trades. Certain sub-scans would just never populate ever again due to overfitting. It always felt one step behind from real-time greatness, no matter how many ‘refinements’ the Professor made to reset the historical data.

The discussion would be incomplete if we didn’t talk about the market itself. Personally, I had career high years in 2020 and 2021.³⁸the CPTs who were in the HSM chat fit the category of seasoned traders who had already established sustained profitability long before joining so they don’t count as “success stories” In the prop world, I knew traders like Clockwork and Eagle made heaps more than I ever made. WTG ³⁹reminder: Western Trading Group, a reference to my old prop firm in Prop Trader Series) made over $100 million, their best year since 2008. Retail traders as a whole were making more than ever via the SPAC craze, NFT boom, and the emergence of meme stocks. Professional investors were making money in a raging bull market led by the Mag7. This little HSM discord that I stumbled upon from years ago seemed to be the only realm of finance where the easy money was happening tomorrow, not today. I can only imagine how discouraging it was to see all of that happening in the background while feeling like you did everything right–working hard, following a well-defined process, backtesting your edge, and trying to match this guru’s supposedly unbeatable system… all just to achieve mediocre results in the greatest daytrader’s market ever.

Wait. Actually, I don’t have to imagine at all because I knew traders who felt exactly that way. I think of Ray, the dock worker who quit in 2019 after a bad spell punctuated by a tough loss on BLNK. I think of Raj, who had an uncanny ability to reverse engineer all of PB’s scans⁴⁰you want to know the funniest shit? Raj also curve-fitted a bunch of a BG scans like the Professor did but in real time, his were way better! there was a stretch in late 2021-early 2022 where I just did nothing but knock down big gappers from Raj’s scan while the Professor’s 5 different BG scans were total garbage. He’s one of the hardest workers I’ve ever met. He went from completely worshipping the HSM 3.0 method to completely disavowing it. I think of Ethan, who had been in the Professor’s room even longer than I had. As a trader with quick scalping tendencies, he couldn’t jive with the HSM 3.0 method that basically made every student a stationary position trader. He moved on. For all of them, the Fantabulous Scan-Traption only delivered wasted time and disillusionment.

Checked Out and Moving On

Once a respectable teacher of markets with time-tested pearls of wisdom to offer all of us, the Professor went full Spiderman villain arc and morphed into Dr. Boring the Mad Scientist–overtinkering in reaction to every small sample of data, complicating trading with nonsensical sub-scans and rule sets, and unable to satisfyingly answer any honest questions about his backtesting methodology. His answers just kind surmounted to… “trust me, it works bro.” Meanwhile, on his public facing @HowStocksMove account, he kept acting like he knew something you didn’t.

I had long checked out from listening to the Professor but now I felt a bit turned off from his tone. I decided it was time to go.

We didn’t trade blows online through a raging argument in front of all his students, resulting in him booting me out of chat and telling me to fuck off. I didn’t type out a long-winded e-mail threatening to leave lest he heed my warning of the errors of his ways. Life occasionally resembles a movie, like when fate would mysteriously work to re-connect us in 2016. But far more often real life is just boring–two guys stop talking and don’t say goodbye.

I hold no ill will. I needed to make a change. Pro tennis players and golfers will often change coaches when the familiar voice no longer resonates with their drive or vision. I’m glad I encountered the Professor in my trading journey–least of all because it made for this good content. The HSM 1.0 era provided me a lot of growth. The HSM 2.0+3.0 era, even though I balked at trading within that style, also gave me something–a stronger sense of who I am as a trader. I would not further waste my time LARP-ing as a systems trader. It’s just not in my DNA.⁴¹just as an aside, I do like data-driven decision making. I read Moneyball when I was a kid and it had a lasting impression on me. I kept track of all my stats as a poker player and then kept track of my trading stats to a greater level of detail than anyone else did at my time at MBC. I just didn’t know what bad data was until this very experience. I’ve also developed the belief that experience and gut instinct are just as much, if not more, a huge asset in trading. It can’t always be data.

Someone with the right skills who knows how to properly data mine and backtest will probably find gold executing a superior version of what the Professor tried to do–I hope that person is reading this and found something good to use. I don’t feel a need to “be right” in any grand argument about discretionary trading vs. systems trading. Don’t involve me in that. I just want to make money and I hope you do too. To the vast majority of the people reading the blog–never forget that we are the underdogs of the market. Whatever you find that works for you, hold onto it for dear life.

Is the Professor killing it today, making millions while flying on his gulfstream? It’s possible he sorted out all the issues. Or maybe he didn’t but was still “approximately right” enough on his edge to achieve scale or maybe even automation. I don’t feel a need to find out because I have moved on. The Fantabulous Scan-traption was his brainchild, not mine, and I wasn’t about to spend endless days and nights trying to perfect it as he would. The journey can often take much longer than anticipated, and perhaps all I witnessed was a glimpse of the challenging years before he finally discovered his holy grail.

A PASSAGE FROM NEW MARKET WIZARDS

Jack Schwager: What is your opinion about optimization?

William Eckhardt, “The Mathematician”, CTA and co-mentor of the 1980s Turtles Trading Experiment:

It’s a valid part of the mechanical trader’s repertoire, but if you don’t use methodological care in
optimization, you’ll get results that are not reproducible.