Get started
INTL BUSINESS MACHINES CORPINTL BUSINESS MACHINES CORPINTL BUSINESS MACHINES CORP

INTL BUSINESS MACHINES CORP

No trades
See on Supercharts
Overview Financials News
Ideas
Discussions Technicals Forecast Seasonals Bonds

IBM trade ideas

AI vs Quantum Computing AI? Pfft, please. Quantum computing is where its at. The AI generation is here. Stocks are pumping because of it, people are crazed because of it and its going to “revolutionize” the world as we know it. How precisely it will change the world, well I could fathom some guesses. But the question is how will it change trading? And the answer, which is entirely speculative but based somewhat on research, is kind of surprising. It won’t. But something will. And that something is quantum computing. Let me explain to you the likely reasons that AI is not going to “revolutionize” trading in the sense most people, I think, feel it will, and how quantum computing will. And keep in mind this is coming from someone who relies 100% on computing, algorithms and, to a lesser extent, natural language models to program and develop trading models. All of my trades are executed based on statistical models that were programmed using excessive amounts of computing resources and, in some cases, natural language models. At the end of the day, AI will not revolutionize trading as AI is only as smart as natural law and mathematical theories. The assumption that AI is going to revolutionize trading is contingent on the fact that there is something that AI can do that people, algorithms or computers themselves cannot. And the reality is, there is nothing that AI can achieve that is not already known. Market makers and financial institutions make millions if not billions each year based on statistical models and computational algorithms that execute these models in trading. AI is not going to change that. The information is already known. The question isn’t “what in the data don’t we know”, the question is “how fast can we mine this data to modify our models to adapt to changing market conditions”. On average, for me to program a statistical model of a stock, I can do a haphazard one in a bout 30 minutes. This will not be a perfect model but it will be a functional model that would have anywhere from 70% to 85% accuracy. Theoretically, AI would be able to execute the same task in probably maybe half the time as me, but this is essentially only because of cutting the manual labour. The computing time required may actually be a bit longer. Why? Because the computing power needed is vastly greater. We need the computing process to run the AI model IN ADDITION TO the computing power needed to power the statistical analyses of the data. For me, on average, it takes my Microsoft Surfacebook Pro with an AMD Ryzen 5 2.30 GHz and 8 GBs of RAM about 25 to 30 seconds to processes the creation of a basic model when I incorporate only basic trading data. If I bring in volume data and have it calculate distributions and price accumulations within bell curves and have it run multiple probability simulations, the processing time can range from 1.5 minutes to hours, depending on the size of the dataset. Que the quantum computer. Theoretically, a quantum computer could perform this task in under 5 seconds. How? Well its complicated, but essentially quantum computers operate on something called a “Qubit” vs a normal computer which operates on Bits (i.e. bits per second). Bits are sequential. So the computer will execute each task sequentially, which takes time. Qubits are able to execute multiple tasks simultaneously. Hence reducing the processing time. AI is not exempt from the bits rule. It processes things sequentially and has to also make room for the fact that it is processing a natural language model in addition to other models required for statistical computing, thus increasing the processing demand and reducing the speed and efficiency of processing. The major ability of large financial institutions to make money comes from their ability to processes multiple streams of data, from order flow data, to regression models, to probability distribution modelling, to running simulations such as monte carlo simulations of potential outcomes and finding which model fits based with the current price action. All of this requires incredible processing power, processing power that not even AI is going to be able to manage. However, quantum computers, on the other hand, are ideal candidates for processing these large amounts of information efficiently, accurately and, most importantly, quickly. Quantum computing will be able to track order-flow, create simulations based on order flow and current price action and predict and fit the most likely outcome of price trajectory based on real time data. If I were able to have order flow access, like actual order flow access that is only available to financial institutions and exchanges, and connect this to my computer, draw monte carlo simulations as well as instantaneous regression models based on this and historical behaviour and have my computer model and fit the current PA and order flow information to the most reliable simulation within the confines of a broader regression model, I would be accurate probably 95% of the time if I had to haphazard a guess. The issue is, I do not and not even financial institutions have the computing power required for such a thing. Theoretically there could be over 5,000 different likely models. Computing all those models and fitting the PA to each of them would require … that’s right.. a quantum computer. That said, AI has been found in the research to serve a unique purpose in terms of trading and financial information. And that is in processing fundamental and news related data. Research has already demonstrated positive outcomes from using AI models to interpret fundamental and news related data on stocks. Goldman sachs has invested huge amounts of money investing in AI research which, among other academic and financial institutions, have lead to very positive findings in AI’s ability to interpret how fundamental, economic and earnings data will impact stock trajectory (if you are interested, the title of a published article that researched this is titled “A comparative study on effect of news sentiment on stock price prediction with deep learning architecture”). The reality is, both AI and quantum computing will have profound benefits to those who can afford them. The richer will continue to get richer. But the most profound impact on direct trading will come, in my opinion, from quantum computing. Because the inherent process of stock price prediction is related not so much to the ability of AI to processes language data, but the ability of a computer to processes price action and order flow data and model it accordingly. This does not require AI. It just requires algorithms. It probably won’t surprise you to learn that the biggest investors in quantum computing are … big financial institutions. And if you look at the top investors for quantum computing, its predominately all financial institutions. If you look at the top investors in AI, financial institutions don’t even comprise half the list. Yes there are some financial institutions, but they are not the rule. There is more diverse interest in AI than in quantum computing, which, I believe, speaks volumes to the implications of quantum computing vs AI in the financial sector. Me personally, if I had to pick between owning a fully capable AI model vs a quantum computer, I would pick quantum computer. Because I know my process works but it is limited by the limited computing power I have. A quantum computer would be my ticket to richness. AI would be meh. Thanks for reading everyone and as always, safe trades! And I also just want to quickly give a shout out to IBM. IBM currently leads the world in quantum computing and initially lead the world in AI (google IBM Watson if you don't know), and yet their stock continues to under-perform the rest of the market.
NYSE:IBM
by Steversteves
3
IBM: Processing… 💻IBM is still processing wave b in red – the second step of a three part downwards movement, which should soon lead to the low of wave B in turquoise. As soon as wave B in turquoise is complete, the share should turn upwards to climb above the resistance at $139.47. There, IBM should conclude wave 2 in green before a significant downwards movement should take hold. However, there is a 35% chance that IBM could rise above $139.47 directly to develop a new top of wave alt.A in turquoise first before turning downwards again.
NYSE:IBM
by MarketIntel
Mega running triangle in the makingI inadvertently stumbled upon the chart of IBM last night and I was intrigued with the pattern that has developed over the past thirty plus years since 1999 (the dot.com crisis). Economists and market analysts have been talking about the US economy undergoing a recession of late because of the interest rate hike and the sticky inflation, with the US stock market showing signs of bearishness. I am not on the side of the bearish camp and I believe that we are currently undergoing a correction which will end in the foreseeable future. The financial stocks are a good harbinger for the US economy and I believe the correction of the financial stocks is going to end soon. I have been using goldman sachs as a proxy. I will post my analysis of GS when the right opportunity arises. Back to IBM. I reckon it is on the path of completing the wave E of the 'mega' running triangle that has developed over the 34 years as depicted on the weekly chart. Let me expound on it. Wave A : W -X - Y formation. W is a flat, followed by X, and Y is a zigzgg. Wave B : The longest wave in the triangle that explains why we are getting a running triangle. I will breakdown wave B into a W-X-Y formation. Wave W is a abc pattern. I will label this as a elongated flat. 'a' wave is a leading diagonal follow by a steep descent, wave b, and a five waves 'c' wave. Wave X is a regular flat pattern followed by Wave Y, a triangle. It may be difficult to see the triangle legibly on the weekly chart, but can be viewed visibly on the daily chart. Wave B is labelled as a double combination wave. Elongated flat (W)-(X)-triangle (Y). Wave C : It is a classic zigzig abc pattern. 'a' wave is a five wave impulse pattern with third wave extension. 'c' wave is a expanding ending diagonal. Wave D : It has sufficient waves to label it as a W-X-Y pattern, but corrective waves are tricky. A decisive breakdown below 125 will give a good indication that wave D is completed and that we are on the path to wave E Wave A, B,C and D ( white label) are all but corrective three waves. Once wave E is completed in the region of 100. IBM will be riding on its impulse wave C (orange label) to a new high. IMHO.
NYSE:IBM
by brown_maverick
Crooked W // Bearish CypherThis pattern is called bearish as it looks like a crooked W and the 1st leg is down. The 4th leg is an up leg and terminates near the .786. It is only bearish if you get stuck at the top of that 4th leg since this pattern reverses there. There is money to be made or a long if you catch a bearish harmonic pattern close to the bottom of the 3rd leg. Valley 2 is lower than valley 1 in the bearish version. In the Bullish Cypher, peak 2 will be higher than peak 1. This pattern shares some characteristics with the Shark pattern. The Shark seems to be more common than Cyphers are. No recommendation. There are a lot of red candles in a row representing a large amount of selling pressure. This could lead to a pull UP soon. It is not quite oversold with RSI reading 36.6
NYSE:IBM
by lauralea