Hi folks, I've been pondering a first purchase of a used mini excavator. I usually build some kind of predictive model to see if an offer is a good price, given several factors. Here's my first attempt: https://docs.google.com/spreadsheets/d/1dv4wSVJwAwMhQMKDU8zj4U2-ygMrUs0AVMjZIIGaeyQ/edit#gid=0 I've got the accuracy of the model to within $8500 of asking price on average, but I think I need some more features to improve the accuracy; I'm aiming for around $5000 average error. So my question to you guys, #1 One feature in particular I was some adjustment factor for the brand's reputation. All else being equal, just knowing the brand of the machine, which would you choose? If you'd indulge me, could you give a top 3 or top 5 best in your opinion, e.g. 1. takeuchi, 2. kubota, 3. case, etc? In an ideal world, there'd be plentiful reviews of each mini model, but that would be asking a lot. #2. Am I missing some major factors in my model, apart from brand? I was thinking of adding 'total hydraulic flow' as a feature, though I would expect it correlates pretty well with the gross power. #3 Is there a source of consumer reports for mini exes? I've looked and looked. #4 Is there a source of tabulated data for mini exes? What I found on this forum was a couple of excellent posts by StumpyWally - e.g. HeavyEquipmentForums These are a little out of date though. Any input on these would be a big help, and maybe this model will help others - they could add their own lines to it and assess the value of an offer or their own machine. Thank you all
Welcome to the Forums LB! Best of luck on your search.
Wow! You have put some time into that spread sheet. It could have some value. However there are many aspects that should be considered, and some of those are subjective. Like Ergonomics, feel of the hydraulics, speed, mulitfunction ability, serviceability, dealer access, resale and so forth. Some of your decision should be decided on your seat time with it. Your spread sheet has a lot of different sizes from a 35 to a 90. I think the first goal would be to decide what size of an excavator you need. Once you make that decision, it reduces a lot of clutter.
Thank you kind sir. Thanks for those thoughts K, I will keep all of those in mind. Dealer access and supply of affordable parts is a big one. My assumption is that continental brands will have easier supply of parts but I could be wrong. I will have a welder shop to fix some things. Serviceability is something I'd like to get a feel for from user reviews -- I guess that means reading this forum thoroughly. I hear you... I'm looking for a 40-55 I believe, just from watching Youtube videos. This one for example, I really like the speed of that Yanmar and it's a Vio 55. I had rented a lighter mini before and didn't like how it would begin to tip over with an extended bucket of crushed rock. For the purpose of data analysis, I do need a bit of a range so as to fit the variables and be able to extrapolate. As for resale value, that's a large part of the exercise, to determine how much these machines depreciate per hour usage, per year, etc. This way I can compare it with renting prices. As it stands, ownership looks favourable. $3-4 in depreciation per hour used is nothing. You'd spend more than that per hour in fuel these days.
Takeuchi / Bota / Yanmar / IHI which there getting harder to find , which is now Kato . The new John Deeres are Wacker Nuesom in the small stuff . @ least they were last time I checked.
I'm intrigued by your spread sheet. I don't understand what you are doing with the three value columns. Can you explain what you are doing there? I used predictive analysis sort of like what you are doing for a Cat dealer for years. As far as trying to assign and apply objective data to subjective feelings or claimed machine performance statistics, it's impossible. What you can do is apply market data per machine model over time to show a trend and go from there. Machinery Trader has a free fully searchable data base with auction sales and retail asking prices. You can delineate by model, year of manufacture, location and even price sold or asked. You can also set date ranges on the auction data and track values as the machines age.
Hi Lynxie: I'm not a professional heavy equipment operator. I lucked out by purchasing a U55-4 Kubota (2016) with 1600 hours on it for under $40k. The right place at the right time. My key learning from the pros on this site and about 400 hours on the machine now: 1) need close reliable service, 2) subjective aspects include ergonomics, look and feel, etc..., and 3) key attributes for me (enclosed cab with AC, angle blade, power to take down 60' Oak trees with tap root, rubber tracks because I have a lot of asphalt driveway to traverse on the 18 acre property, reputation for reliability, thumb). I rented this machine twice from a local guy and then just made a cash offer when I heard he was running lighter equipment as rentals (easier to haul around). He took the money. I've never driven another excavator. I do have a backhoe/tractor so I set the excavator controls to SAE (like my bachhoe) instead of ISO and almost no learning curve on the controls...except manipulating a tracked vehicle. 4) Rent one or more machines before buying. I think that experience will help you calibrate a lot of dimensions: size, weight, controls, what really has the value for your particular investment, etc... You may rent a machine and hate it and know immediately that it is too light weight for your application. I'm not a Kubota nut I just happened to acquire this machine--and I think it is a pretty good one. General observations about my specific machine: 1) the foot controls are off center a bit and not ergonomically situated. If I were operating this as a full-time gig I might be annoyed, 2) The cab is a bit tight...I'm 6'2" and 185 lbs. A bigger guy might have issues with cab size. That said, the AC and heat are terrific, 3) Every time I idle the machine and get out of the cab I must re-engage the thumb control on the panel, 4) some of the boom and stick hoses could be more protected from tree limbs/brush clearing. I've knocked a fitting loose a couple of times. This might be more of a skill issue (or lack thereof) than a machine design issue, 5) The KX057-4 is the non zero turn radius brother of this machine. I might have preferred that machine since I'm generally operating in open space clearing land. The KX057 is a bit heavier but with the same general power. 5) Plenty of power for my specific application which is clearing rocky land and trees under 2' in diameter not more than 60' in height. If I were to buy again I'd love a 6-way blade and would seriously check out the Takeuchi. The Takeuchi may have better cab ergonomics, 6) I do have a local Kubota dealer that is very responsive and competent. Kubota's are pervasive in my area so getting parts, etc... no big deal, 7) I have three buckets: 18", 24" & 36". Very easy to switch out, 8) this machine, albeit low hours, has been bulletproof. Never misses a beat. The diesel regeneration has only occurred 4 times and never interfered with my work schedule, and 9) Biggest complaint? Mice in the cab. I may now have solved that issue using an electronic ultrasound device in the engine compartment and peppermint oil on the cab floor. Best in your search, A
Hey Allan, much appreciated giving the in-depth info above. I'll get back to that later since I'm at work. One of them is the prediction of what the average market value is for that excavator. It takes a 'sold' variable into consideration, because sellers ask high and buyers bid low - so there's usually a delta between sold and unsold prices. You cannot directly use unsold goods as evidence for a market price, because nobody has agreed to buy it at that price. I could ask $1m for a crappy machine, but that alone gives no info about what I'll get for it. However, there's a whole lot of data going to waste if you cannot use unsold data. So in my model I have an adjustment factor which accounts for the average delta between unsold vs sold goods. Subtracting that delta from the asking price gives a decent estimate of what it WILL sell for, given time. The bid price informs me as to what offer to make, and also lets me use ads with no sale price as evidence for market value. The 'net value' columns reflect the two different prices - though, for items that are sold, the values are equal since bid price = ask price. My approach was going to be this: say I can get a set of user rankings of brands (Jeff says Cat > Hitachi > Kobelco, Amy says NH > Doosan > Hitachi, ...), I can then take each pair in the list and apply a rating algorithm like ELO in chess - the winner decided by their preference rank. After processing through everyone's rankings you'd have an overall ranking of brands by their rating. This rating I could add to the model as a feature. Over and above that, you're right there's little you can do to quantify things. My attitude is that you use empirical data for a far as you can and use your intuitive judgement for the last mile.
So how do you come to your adjustment factor? Is it a percentage of something, a formula that includes a large sample or maybe a bell curve with a standard deviation. As far as using people's stated opinions of brands, I go back to the markets and see how many machines are offered and sold within a stated period of time, say six months, a year or five years. You can't get enough people in a survey in this industry that would provide you honest and informed answers with the information you seek. If you find 56 Cat 305.5E2 excavators sold at auction in the last year and 55 Bobcat E55, 25 Komatsu PC55MR2, 80 Takeuchi TB250-2, and no Case CX57 units. Now you have a starting point to dig into. Why no Case minis? Why more Takeuchi machines than Cat or Komatsu. Why so many Bobcat machines. One might conclude price is maybe the cause go back to the markets and check. Average auction on the Cat is $40,801, Bobcat $42,720, Komatsu $25,566, Takeuchi is $28,290. So a used Bobcat brings more at auction that a used Cat? Maybe need another look. Checking a little deeper it looks like the Bobcats are mostly late model low hour machines probably out of rental fleets. So let's delineate that a little more. Let's limit the years to 2014 to 2016 for each. Now the Cat machine average auction is $36,274. The Bobcat is still amazing at $38,900. So now we should look at the advertised hours on the machines to see what kind of use that are seeing in six to eight years. The highest hours I see on the Bobcat is 3,500. I see one Cat showing 3,900. At this point it looks like both the Cat and the Bobcat machines are not being utilized. No hours means no revenue and payment still have to be made. So bad investment in my eyes. How about the Taki and the Kommy. I see one Taki at 4,765 and I find one Komatsu with 6,321 hours. The only real conclusion to be made is you can lose your shirt on a mini unless you have plenty of work to pay it off quick. You see how deep the pool is on buying iron. I think you are making a good effort, but the buying of any equipment will be based on what you can afford, what is available and where it is at. Whether or not it is a good buy will depend on what the buyer knows about the equipment type that he is looking at to weed out the black holes to poor money down.
It's a percentage, and like the other parameters it's fit to the data using gradient descent with mean-squared error as the loss function. I tried this adjustment factor modelled as a constant, but got greater error overall so a percentage works best. It's simplistic for sure, but I'm wary of increasing model complexity with such low data amounts - way too easy to overfit and undergeneralize. It sounds like your approach is to see how popular machines are, as a proxy for measuring quality which itself would be difficult. For new machines, you'd have distorting factors like the amount and competence of their advertising, deal-making with distributors to keep other players out, and so on. For auctions though, popularity, and prices that exceed the market average, give you more of an honest signal of the quality since the brand doesn't put any effort into selling those. Good call! Also that's an insightful thought about low-hour (i.e. rental) machines - why are some models getting underutilized? There could be a clue in the data. As always it'll be noisy - renters often won't know much about different mini-ex models, as was the case for me. It was just "I need a job done...oh this guy is renting, let's give him a call"
You can forget about buying a Kato / IHI new from Kentucky within probably 10 months . I just called / inquired to swing by and purchase a new CTL and they are completely wiped out of CTL's and xcavators .
Hm. I am very curious what the 1-2 year future will look like. Will prices come down because of high interest rates and payments, or will they still go up because of high inflation? :s
Prices for new have never gone down. Only the rate of increase has changed. The price of used equipment is tied to the price of new iron. It used to be new iron prices only changed once a year. Now most all manufacturers raise prices twice a year. I have two books on the subject of machine condition and values available on Amazon. They are available at the link below: https://www.amazon.com/~/e/B0742Q36JL
Yep this is what I do to value our used iron, super easy and very accurate. Got to MT like John said and find the machine you want to research value on. Go to the For Sale and Auction Results tab. The Auction Results are a direct reflection of the market. Take the average For Sale price with high and low removed and do the same for Auction Results. Add those 2 numbers together and divide by 2, very complex. This value will be very close to what the machine is worth for the age and hours of the machine. The more machines in the For Sale and Auction Results the more accurate the value will be. Bought and sold a lot of iron using this method.
Consider condition. Does it have slop? (Play) how does it look? As far as brands I like: Takeuchi by far #1 for performance. Catapillar #2 Case or sany #3 The natural gas company in my state has switched to exclusively Takeuchi Excavators. They said because it was the best machine for their needs.
Welcome to the forums Drjk! Interesting our natural gas company here just switched from Yanmar to Cat.
As someone who loves having a set value for good ideas vs bad ideas and where to put my money. If your goal is to buy a used compact excavator go talk to local companies that use them and see what dealer gives the best support. Numbers don't help much if everyone else in you country gets great service from one manufacturer but your local branch is useless. When your machine goes down are you always going to take it into the shop or try to fix it yourself? Will the dealer send you schematics or walk you through "quick fixes" over the phone? Those questions are much more important to a machine owner. Your spreadsheet is more in line with a rental company or fleet management goal.
I just traded in a kx040 in ky, dealer might still have it. Great condition.
It took a bit of work, but adding a factor in the model for each brand increased the accuracy quite a bit. This is the brand premium it produced by fitting the data, with some regularization. CAEL has by far the lowest brand premium, as they only have for sale this severely discounted, and frankly suspicious 2022 6.2 tonne model - https://www.capitalautoequipment.com/product/excavator-6-2-ton-with-yanmar-engine/ JD and Cat turn out to sell at a premium perhaps due to marketing but also local dealers. Overall I'm quite happy with the model. It's average error to market price is about 13%. When optimized for absolute price difference, it was about $7600 off on average. The prediction vs actual price has a correlation of 0.93, where any individual feature only has relatively weak correlations of 0.66 (condition) 0.61 (manufacture year) 0.55 (weight) 0.51 (power) 0.43 (sell year) -0.36 (hours) Meaning, if you knew just one piece of information, for example the condition, you'd have dreadful predictions. There was a video I watched where a general contractor claimed that Mini's depreciate not by age, but by how many hours it's used - i.e. don't let the engine idle 'cause it'll cost you. Well, this puts the lie to that... hours used is a very poor measure. Condition matters most, then age, then its scale (weight and power), then oddly, sale year. All of those are better predictors of price than 'hours'. The model suggests the best deals by absolute discount from fair market value, are: Asking 97,500, assessed at 115k - A 2019 Volvo EC60E, 152 hours, 5/5 condition: https://www.ritchielist.com/constru...-60/5cbedead-c645-4d08-bdb5-a8fa4e6506db.html Asking 55,000, assessed at 70k - a 2013 KX040, 860 hours, 5/5 condition (ad taken down ) Asking 70,000, assessed at 84k - a 2021 6200kg CAEL, 2 hours, 5/5 condition. This even though CAEL is discounted heavily for its weird unknown brand. For comparison if it were a Kubota same specs, it'd be assessed at about 102k
It's definitely a good point. I'd like to be able to just order parts online, watch some videos, and DIY. I do weld and know my way around engines, but... not necessarily looking forward to that as a part time job. I don't have a feel for how much maintenance these machines are, so that could be an unrealistic time burden.
Lynxie: You have very advanced statistical modeling skills. I could have used you in my consulting business! Just remember the saying, "When all else fails, manipulate the data." : ) That said, I think you've completed the general analysis. This work should help you narrow the field of available excavators (machines that you have an interest in pursuing. Now, per the advice of the "real experienced guys" on this site, go rent a few machines. Talk to dealers and locals that are using excavators that meet your parameters. I'd be interested to hear how your real-life experience combined with this detailed analysis affects your purchasing decision. Best, A
Thanks very much Allan - always nice to hear when you've put some effort into something. Yeah I agree, homework done -- now for the labwork.
Is all your decisions this tortured?
Yes
Zillow tried to do the same thing with houses. In the end, they shut down that part of the company and laid off all those employees. Unfortunately, you can only determine a range of value using statistics. Statistics also have to have a trusted and reliable source. No mater how you look at it, purchase and sale prices are still a matter of feel based on time in the industry.
I would rank CAT in first place for no other reason than I believe the have one of the best parts systems. But I am biased.
I'll chime in with my usual comments on this issue, dealer support should play a huge factor in your choice. How quickly can your dealer get you back to work when something goes wrong? How well equipped is their facility? I'm very fortunate to have a Caterpillar dealer 20 minutes away, they have set the model for machine support both with parts availability and service facilities. I also have Deere and Komatsu within 30 minutes. I have a Cat 304E2 at work that does a nice job, my only disappointment has been starved hydraulics when running a brush head with the auxiliary hydraulics, I can't move the boom or dipper if the brush head motor is powered. If you're planning a lot of this type of work I'd look for a machine that has different priorities in the hydraulics or a dedicated pump for the auxiliary if such a machine exists.
First off apologies, didn't see the response. When I worked at a deere dealer we were very big on helping equipment owners decide if it was worth bringing in simple troubleshooting over the phone was always an option. But we also had customers who had plenty of money but not much time and they would drop off their equipment for any thing. I don't remember any of JD mini ex's having too many major problems but lots of leaking O-rings with age and heat, compact means less air flow, and busted or worn hoses. None were overly complex but some times routing a hose can be rather frustrating.
LB - spent an hour reading your submission and looking at your spreadsheet. Great piece of work. Very likely you will be able to make money using any of the machines, but agree you need to make sure you get your first machine at the right price with lots of life left in it. Interesting comment wrt hours not being one or two in the list of important attributes. I have a few pieces of construction equipment ( 2 cranes, Boom Lift etc ) and am about to buy my first excavator, your spreadsheet affirmed that I paid the right price for it.
Any update?
Thanks for the info Steve. Thanks Z Nice! Saving up cash...don't like debts.
Hello Lynxie, It sounds like you are where I was about 18 months ago. I spent a lot of time doing research on brands and what and how to buy. I wanted a reliable solid used machine. I considered all the usual suspects and went and looked at several. There were no great deals out there but by doing research I knew the market and I was ready when my machine came on the market. I bought a 2014 Deere 35G with a thumb and three buckets with 1,200 hours on it. I was open minded to several brands and ended up closing on my 35G. I love it. As others have stated close availability of parts and service is important. I looked at sold pricing from Machinery Trader and Ritchie Brothers to help me zero in on market pricing and it turned out to be pretty accurate. I learned that you have to know the market and be ready to pounce when a machine you like hits the market. After you buy it, don't look back just go forward. I have attached my spread sheet which is kind of like yours (but not as sophisticated as yours). It helped me be confident in pricing when my machine came up. If you have any questions about my buying process or my machine just ask. Good luck.
Here we are 9 months after your post, looks like up 30% between May and now. IF YOU CAN GET ONE!!!
You're sure on it with the analytics. Some great data gathering and analysis. Coming from a former numbers guy, turned into operator/contractor purchasing the right machine goes quite a ways beyond just those numbers. Pretty much all of the mainstream brands are good brands. A machines price or market value is not necessairly reflection of the brands quality, reliability etc. Some brands simply have much more overhead to pay for than others. Just make sure you can get parts easily/ Know what other buckets or attachments you may want for the machine. Lugging makes a difference. For example it's pretty easy to find used attachments at a freaction of the cost of new if you have a wedge coupler. if you end up with a wain roy or a bobcat x change, good luck. If I was in your shoes, I'd look at power - go either yanmar or kubota. Wouldn't put as much weight on actual brands as they all have their pros and cons I'd stay away from auction equipment and try and get an o/o machine with service records. I know guys that literally buy junk machines, get them to work and flip them in auctions. Prices will most likely keep coming down as rates go up. Key is to take your time and the right machine will come up.