LC

Spek ;D

Whats the reason for the switch? You see more upside with PayPal? Long term yes. Short term also capitulating to any bearish ideas. The original idea was to invest in some recession-resistant dividend paying companies with at least some pricing power. But growth has totally outperformed this idea. Mostly I think I need to move towards a more passive (indexing) option, at least for the near term (3-5 yrs I'd guess). Over the last year I haven't really had the time/energy/pleasure to actually do any investment research, and the results reflect it. Even worse, I haven't really cared... The reality is given my situation (age, net worth) my ROI is higher concentrating on my career than on my portfolio (ROI both in terms of amount and volatility of cash inflows). Something about whole-assing one thing vs. half-assing two things ;D ;D

Sold Verizon, bought Paypal

They are absolutely worth it. Silicon valley investors are a dime a dozen and add little differentiated value. Quality engineers on the other hand are a rare commodity. If your friends/family are truly the latter they should be out on the west coast getting paid.

Take it to the extreme: you only review and know the answer to 1 question. Is your skill level still very high?

But is this luck? To me this seems like pure skill. By applying enough skill you can guarantee a 100 on this test. No luck is required. Luck is applying zero skill, having the teacher get a better job at the school across the street and having the test cancelled.

Beautiful story! So happy for Chica and Max. Dogs are the best :D

IMHO and in my experience ML techniques are better utilized in the consumer space, particularly consumer credit. These models provide some lift over traditional logistic approaches BUT I would argue it is still requires more time to manage stability concerns. Even forgetting regulator concerns, boosted trees, neural network variants and such are so difficult to understand you do not know whether this lift is attributed towards additional intuitive patterns being exploited, or random chance or some other factor which may not be relevant. Interpretability is a real concern. On the trading side there is obviously much more opportunity, but this is why these models will IMHO have a more difficult time adding value here. One these markets are so picked-over it is ridiculous. Every intelligent statistician wants to design trading models. Two the majority of buyers are sophisticated. Compare the average derivative buyer (i.e. a highly educated bank/fund trader) to the average credit card buyer. Three, securitized financial products have more generally-accepted pricing mechanisms. (BS, Heston, etc.) What does the ML model add here? ML models excel at finding nonlinear patterns, and perhaps you can combine this with a traditional BS approach but you still run into the problem of interpretability. Obviously I am speaking in hindsight here but it has become clear why ML techniques are having a difficult time on the trading side of the business. In this sense, I think a lot of the folks who jumped into this business based on the glamor of combining their ML experience either from a digital or consumer credit space towards a trading landscape, they have found themselves in front of something of a brick wall and are withdrawing their claws and turning heel so to speak. Another area where ML approaches perform very well is in fraud use cases. Fraud techniques change rapidly (remember the old Nigerian Prince scam?) and fraudsters are adept at gaming basic systems. Easily you can see why the ML model excels here: it is highly adaptive and specializes in pattern seeking. As with all modelling you are not always looking for the best short-term fit but you are looking to match a particular methodology's strengths/weaknesses with your problem/use case. One big drawback of ML models is managing overfitting. These models have traditionally shown very real stability problems out-of-sample. Regularized learning intervals is one method used to manage such risk, however this leads to "momentum" as I alluded, as the model is regularly updated to take into account and weigh more heavily the most recent data. Curious what your thoughts are in terms of overfitting and managing stability in your particular use case, or prior use cases where you may have had experience.

Are you talking about QC? No need to answer of course, but NYS has been purchasing from HydroQuebec for decades. http://www.hydroquebec.com/international/en/exports/markets/new-york.html On the topic of human divisiveness, while I too share the concern I would argue that in times of real, in-your-face catastrophes, people do band together. It is human instinct, we increase our odds of survival when we work together. In NYC I saw this first hand with 9/11 and Hurricane Sandy. My hope in humanity is not dead yet ;D

Lightwhale, any thoughts on ML applications either generally or in specific cases? We find they are, in some popular use cases, essentially momentum strategies.

Another interesting thing to note is that, the US electric grid was cobbled together piece-meal. In terms of things like national security and such, apparently this is beneficial in that the lack of inter-connectivity ensures there are little to no critical bottlenecks that could become terrorism targets. In terms of NYC, Con Ed is a mess but it is not really their fault. This was not a modern, planned city. It has evolved over centuries. Look at the history of how the city was built and how these buildings get heat and such. I mean, they are still using steam for temperature control! https://en.wikipedia.org/wiki/New_York_City_steam_system So while it is understandable that old technology will suffer unreliabilities (or not if you look at it from a survivorship bias perspective), it is much more expensive for NYC to lose power, than say Bergen County. In terms of the people living and their health as DTEJ mentioned and also in terms of productivity. Banks, corporations, etc. going down for 24 hrs is much costlier than if Joe's Bakery can't make some pastries for a day. And, residents pay very high city taxes comparatively, partly to prevent this sort of thing. So an investigation is certainly warranted. Perhaps we should involve Bob Mueller? ;D

I am appalled by the lack of taste of some friends on this thread :P Kidding. I could wax poetic about schwartz's for hours but I'll spare everyone. Just eat, drink, laugh and smoke on the streets of montreal. Cannot have a bad time with those ingredients. ;D

Bingo. The tri-state area is not the same "NY" as Flatbush Ave. Yes, hence the:

Thanks for the recommendation! I didn't find a place but I have reservations for Toqué and Damas. Are those ok? I also have reservations for Cafe Boloud and Bar Isabel in Toronto and Riviera in Ottawa. I wanted to try Joe Beef in Montreal but there was no availability online. Where do you recommend to have breakfast? Hi Dan, We ate at toque, pastel , and joe beef (we had some friends who got us in last minute). I thought pastel was best. I am glad you cancelled your reservation at toque as I felt it was underwhelming. The best thing about the meal was the service. Although I will say the wine list looked good. Pastel was hilarious. Head chef kept talking all kinds of shit and was a general mess of a human being. Loved it. Try calling - they may be able to squeeze you in if you are so inclined. Let us know how damas is, it was on our list but we didn't have time (or energy). You can also hit up liverpool house which is the same guys from joe beef and same location. Schwartz's is a necessity of life. Medium sandwich, I go fries a pickle and a cherry coke. Others believe the slaw is necessary, or a non-cherry soda. These poor souls have been misled and you should give them no mind. It;s cash only and there's a line if you go during normal human lunch hours. Nobody has ever said it wasn't worth it. Personally I like to walk the neighborhoods when I visit. Plateau, old port, and griffintown are all incredible. The city, particularly the central area, is very walkable. Old port has tons of construction going on so it is a bit more crowded on the sidewalks. I love walking St laurent, shopping in the boutiques. Mt Royal is also beautiful as cigarbutt mentioned but it's a walk - it depends how active you like to be on your holiday. Two other more affordable/cozy spots are breizh cafe for crepes in the morning, and lola rosa (the one in the mcgill ghetto) for vegetarian lunch and goblets of sangria ;). Just take everything I say with a grain of salt, I prefer a "grittier" experience than most, my nights wind up drinking cheap beer @ biftek (a student dive bar on st laurent) for three hours, eating 2 dollar peanut noodles across the street at 2 am, smoking too many cigarettes, and passing out until noon. Barfly is another gritty dive bar if you like that. Actually there is another nice spot called le majesqitue right across the street from barfly that is yummy. Le distillery (i think its renamed distillery no. 1) is a cool spot for mason-jar cocktails. Again a little grittier. Bar Furco is a nice trendy spot, there are other similar trendy bars east of peel and south of sherbrooke. If I am drunk late and have been wondering I wind up in this area. The french party late. West of peel (again below sherbrooke) you have places like brutopia, this area is a little more anglo-friendly and has more of the pub vibe vs the cocktail vibe when you go further east. Let us know where you go and how you liked it! I keep remembering stuff. Patati patata is great poutine. Beauty's diner for a morning bagel (beauty special) and coffee.

I would look at the e-sports area as well. There was a blog post a few years back on the favorable economics of businesses which run industry conferences. I think it was in relation to Nielson or Reed Elsevier (now RELX). I forget the exact company. But the jist was, organizing these trade conferences was a good business as you get paid up front by vendors and attendees, and it's a recurring item. To that end, if I were to invest here I would invest in the growth of these e-sports event organizers. The thing with gaming is, it's a lot like VC investing. Every year get tons of trash games but a few massive hits. It's difficult to know which will be the hits beforehand. But event organizers are getting paid after-the-fact, once these hit games have been established. Takes the speculation out of it.

If you go to Montreal, have a dinner at Pastel (if budget allows). Best restaurant in the city!

Happy 4th! Happy birthday! Most importantly, enjoy the festivities!

They may be the only humans you can be sure about :D

I don’t think anyone is arguing educators inflate their numbers. But I think the point is that if colleges prices hadn’t been inflated so much, this would be less of an issue. In other words, it is not a root cause but rather a symptom of.

Just make these loans default-able in bankruptcy after 8 years.

Attack on Titan. Incredible show!

The Curry's are good stock (heh) I was pulling for the Warriors (boo injuries!) but it's a great day for Canadian basketball, so congrats Raptors!

sold TFSL, RELX, DIS, bought some freaking Visa (not with all the proceeds, just a tiny position) :o :o

I agree with SD especially regarding the working hobby aspect. Will be a better use of your time. But I'd say take 5 or so % and play with that.

https://qz.com/1623418/index-funds-now-account-for-half-the-us-stock-market/ Investors have put $4.305 trillion into passive US stock market funds as of April 30, only $6 billion shy of the $4.311 trillion overseen by active US equity funds, according to a Morningstar report. This month’s data will very likely show that the passive funds have surpassed those that are actively managed.